Home

PDF - IBM Redbooks

1. VMware PSP 3rd Party PSP Figure 7 40 ESXi Pluggable Storage Architecture PSA ESXi contains default plug ins for most storage devices but vendors can provide their own plug ins to fully utilize any special features of their storage devices While the SATP should not be changed from what ESXi has chosen all three of the following PSPs are supported for the DS8000 gt Fixed All I O is always routed over one of the available paths The user can set one path as the preferred one which is always used If this path fails a failover to one of the remaining paths is done If the preferred path returns a failback is done With PSP Fixed a manual load balancing is possible by choosing different paths for each LUN Most Recently Used MRU Similar as for Fixed all I O is always routed over one of the available paths In case of a failure a failover to another path is done But if the initial path is recovered no failback will happen Manual load balancing is not possible for this This PSP has initially been developed for active passive arrays Round Robin RR This PSP is the only one that evenly uses all available paths I O is routed over all active paths alternating If one path fails it is skipped After its recovery it is used for I O again automatically 7 3 2 Path naming ESXi uses a special scheme to name the paths Each path to a LUN gets a runtime name according the scheme in Figure 7 41
2. vmhba3 C0 11 L2 LUN number Target DS8000 port Channel usually 0 ESXi host HBA port Figure 7 41 Path naming scheme Each port is specified by four parameters gt gt gt ESXi host HBA port The Fibre Channel adapter port on the ESXi host Channel The storage channel Target The DS8000 host adapter port that the path is connected to LUN number The LUN from the DS8000 Chapter 7 VMware vSphere considerations 143 The path list for a volume can be viewed by right clicking the volume in the Storage Adapter view and selecting Manage Paths Paths Runtime Name Target LUN Status vmhba5 C0 T0 L0 50 05 07 63 0a ff c2 9Ff 50 05 07 63 0a 00 42 9F 0 Active 1 0 vmhba5 C0 T1 L0 50 05 07 63 0a ff c2 9f 50 05 07 63 0a 40 42 9F 0 Active 1 0 vmhba4 C0 T1 L0 50 05 07 63 0a ff c2 9f 50 05 07 63 0a 40 42 9F 0 Active 1 0 vmhba4 C0 T0 L0 50 05 07 63 0a ff c2 9f 50 05 07 63 0a 00 42 9F 0 Active 1 0 Figure 7 42 Path information In the example in Figure 7 42 the volume has four paths from two HBA ports vmhba4 and vmhba5 to two different DS8000 ports TO and T1 The Target column provides information about the WWNN and WWPN of the storage device 7 3 3 Multipathing and DS8000 IBM does not provide a specific multipathing plug in for DS8000 Therefore ESXi automatically uses the following default plug ins gt SATP VMW_SATP_ALUA is chosen because the DS8000 is an Asymmet
3. Figure 2 4 Enabling the selectable boot b Select the first boot device entry and press Enter as shown in the upper panel of Figure 2 5 Selectable Boot Primary Boot Port Nane Lun ID Vendor Product IBM 2187988 No device pracant Rev 312 568587636AG8829F electable Boot Settings Enab led Port Name Port ID BCAOFAB Select LUN Selected device supports multiple units LUN Status Supported Supported Supported Not supported Figure 2 5 Select boot device c In the Select Fibre Channel Device panel as shown in the middle panel of Figure 2 5 on page 11 select the DS8000 port that you want to use to access the boot volume Press Enter Chapter 2 Open systems considerations 11 d In the Select LUN panel which shows the available ports with their worldwide port name WWPN as shown in the lower panel of Figure 2 5 on page 11 select the logical unit number LUN that represents the boot volume Then press Enter e Check the boot entry that you just created 11 Optional Configure a second boot entry for example to have redundant links to the boot volume To configure a second HBA complete the following steps a On the main option panel of the utility move the cursor down to the Select Host Adapter entry and press Enter b Select the same DS8000 volume but a separate host port for the redundant connection c Repeat steps 1 on page 9 through 10 on pag
4. Maximum 3 900 IOPS Maximum 140 MBps of sequential throughput Maximum 45 54 MBps of transaction throughput gt A port in 4 GB IOP less Fibre Channel adapter feature number 5749 or 5774 can support these maximums Maximum 15 000 IOPS Maximum 310 MBps of sequential throughput Maximum 250 MBps of transaction throughput gt A port in 8 GB IOP less adapter feature number 5735 can support these maximums Maximum 16 000 IOPS Maximum 400 MBps of sequential throughput Maximum 260 MBps of transaction throughput When sizing the number of adapters for an IBM i workload consider the guidelines by IOPS per port by disk capacity per ports and by MBps per ports which are shown in Table 11 6 The guidelines by IOPS and by MBps are to be used for the peak of a workload because they consider 70 utilization of adapters per ports The guideline for the disk capacity per port is based on the number of IOPS that one port can sustain at 70 utilization on the access density AD of an IBM i workload and on the preferred maximum 40 utilization of a LUN Tip Access density measured in IOPS GB is the ratio that results from dividing the average lOs sec by the occupied disk space These values can be obtained from the IBM i system component and resource interval performance reports Based on experiences with IBM i workloads it is assumed that AD 1 5 I OPS GB for the sizing guideline The guideline
5. 0 000 eee 283 13 3 3 LUN Masking renere eda ee easiest eb ae cee 283 13 3 4 Double striping issue 1 0 eee 284 13 3 5 SVC publications 0 00 0000 cae 285 Related publications 00 eee 287 IBM Redbooks publications e resia ra nara E AEE eee 287 Other publications n uaaa aa 287 Online resources ununun anana 288 Hep irom IBM sas aitaa a ae Bie ia a A gh Seat dh Ae 288 viii IBM System Storage DS8000 Host Attachment and Interoperability Notices This information was developed for products and services offered in the U S A IBM may not offer the products services or features discussed in this document in other countries Consult your local IBM representative for information on the products and services currently available in your area Any reference to an IBM product program or service is not intended to state or imply that only that IBM product program or service may be used Any functionally equivalent product program or service that does not infringe any IBM intellectual property right may be used instead However it is the user s responsibility to evaluate and verify the operation of any non IBM product program or service IBM may have patents or pending patent applications covering subject matter described in this document The furnishing of this document does not give you any license to these patents You can send license inquiries in writing to IBM Director of Licensing IBM
6. Disk Pools and click the disk pool that you have just created Environment My Connections My Connections as Basic Operations Aj Name _ Release Description _ H E Work Management ff 9 155 57 42 v7rim0 ITSOi E B Configuration and Service fi 29 155 70 14 v6rim0 Mainz I6PFE System Values fi 9 155 70 15 vSr4mO AS400PFE Sf History Log fi 92 155 70 32 v6rim0 iSPFE bri iil 9 155 70 40 v7rim0 Manage this system Miane z il MceiSli mai v5r4m0 Manage this system 74 Communications System Adapters ha LAN Resources amp Workstation Resources Processor Information i Cryptography Resources g Optical Units H Q Disk Units All Disk Units Ej i By Location E Disk Pools g Disk Pool 1 D Disk Pool Groups aPE Darin Cate A i ff Add a connection Add a connection il Change all of your system passwords Install plug ins Figure 11 26 New logical volumes shown in System i Navigator 234 IBM System Storage DS8000 Host Attachment and Interoperability 11 7 Booting from SAN Traditionally IBM i and formerly System i hosts required the use of an internal disk as a boot drive or load source unit LSU or LS The boot from SAN support was added in i5 OS V5R3M5 and later 11 7 1 Requirements for boot from SAN You can implement the external load source unit with Power5 Power6 Power7 servers and with Power6 and Power7 technology based Blade servers You can also use it with System p s
7. Example 6 1 Virtual SCSI disks p6 570 lparl3 1sscsi 0 0 1 0 disk AIX VDASD 0001 dev sda 0 0 2 0 disk AIX VDASD 0001 dev sdb The SCSI vendor ID is AIX and the device model is VDASD Apart from that they are treated as any other SCSI disk If you run a redundant VIOS setup on a machine the virtual disks can be attached through both servers The virtual disks will then show up twice and must be managed by DM Multipath to ensure data integrity and proper path handling Virtual Fibre Channel adapters through NPIV IBM PowerVM the hypervisor of the IBM Power machine can use the NPIV capabilities of modern SANs and Fibre Channel HBAs to provide virtual HBAs for the LPARs The mapping of these to the LPARs is again done by the VIOS Virtual HBAs register to the SAN with their own WWPNs To the DS8000 they look exactly like physical HBAs You can create host connections for them and map volumes This can result in easier more streamlined storage management and better isolation of the logical partition LPAR in an IBM Power machine Linux on IBM Power distributions come with a kernel module for the virtual HBA which is called ibmvfc After loaded it presents the virtual HBA to the Linux operating system as though it were a real Fibre Channel HBA DS8000 volumes that are attached to the virtual HBA appear exactly the same way as though they where connected through a physical adapter Example 6 2 lists volumes mapped through NPIV v
8. gt Setting up and configuring multipath I O gt Setting up a system to boot from multipath devices gt Combining multipathing with LVM and Linux software RAID This guide can be found at the following website http www suse com documentation slesl1l stor_admin page documentation sles11 st or_admin data bookinfo html Chapter 6 Linux considerations 83 IBM wiki for Linux on the Power Architecture system IBM maintains a wiki that contains information such as a discussion forum announcements and technical articles about Linux on IBM Power Architecture systems This wiki can be found at the following website https www ibm com developerworks wikis display LinuxP Home Fibre Channel Protocol for Linux and z VM on IBM System z The IBM Redpaper publication Getting Started with zSeries Fibre Channel Protocol REDP 0205 provides comprehensive information about storage attachment with Fibre Channel for IBM z VM and Linux on z VM It covers the following topics General Fibre Channel Protocol FCP concepts Setting up and using FCP with z VM and Linux FCP naming and addressing schemes FCP devices in the 2 6 Linux kernel N Port ID virtualization FCP security topics vvvvvy You can find this Redpaper publication at the following website http www redbooks ibm com abstracts sg247266 html Other sources of information Additional information from IBM and Linux distributors about working with the Linux platform
9. jostat nE c2t1d1 Soft Errors 0 Hard Errors 0 Transport Errors 0 Vendor IBM Product 2107900 Revision 212 Serial No 75207814206 Size 10 74GB lt 10737418240 Bytes gt Media Error 0 Device Not Ready 0 No Device 0 Recoverable 0 Illegal Request 0 Predictive Failure Analysis 0 c2t1d0 Soft Errors 0 Hard Errors 0 Transport Errors 0 Vendor IBM Product 2107900 Revision 212 Serial No 75207814205 Size 10 74GB lt 10737418240 Bytes gt Media Error 0 Device Not Ready 0 No Device 0 Recoverable 0 Illegal Request 0 Predictive Failure Analysis 0 c3t1d1 Soft Errors 0 Hard Errors 0 Transport Errors 0 Vendor IBM Product 2107900 Revision 212 Serial No 75207814206 Size 10 74GB lt 10737418240 Bytes gt Media Error 0 Device Not Ready 0 No Device 0 Recoverable 0 Illegal Request 0 Predictive Failure Analysis 0 c3t1d0 Soft Errors 0 Hard Errors 0 Transport Errors 0 Vendor IBM Product 2107900 Revision 212 Serial No 75207814205 Size 10 74GB lt 10737418240 Bytes gt Media Error 0 Device Not Ready 0 No Device 0 Recoverable 0 Illegal Request 0 Predictive Failure Analysis 0 9 4 1 Working with IBM System Storage Multipath SDD SDD is available from your local IBM support team or it can be downloaded from the Internet Both the SDD software and supporting documentation are available on this website http www ibm com support entry portal Downloads Hardware System_Storage Storage_ software Other_soft
10. online invalid cl0t6d0 auto none online invalid cl0t6d1 auto none online invalid cl0t6d2 auto none online invalid vxdiskadm Volume Manager Support Operations Menu VolumeManager Disk 1 Add or initialize one or more disks 2 Remove a disk 3 Remove a disk for replacement 4 Replace a failed or removed disk 5 Mirror volumes on a disk 6 Move volumes from a disk 7 Enable access to import a disk group 8 Remove access to deport a disk group 9 Enable online a disk device 10 Disable offline a disk device 11 Mark a disk as a spare for a disk group 12 Turn off the spare flag on a disk 13 Remove deport and destroy a disk group 14 Unrelocate subdisks back to a disk 15 Exclude a disk from hot relocation use 16 Make a disk available for hot relocation use 17 Prevent multipathing Suppress devices from VxVM s view 18 Allow multipathing Unsuppress devices from VxVM s view 19 List currently suppressed non multipathed devices 20 Change the disk naming scheme Chapter 10 HP UX considerations 191 21 Change Display the default disk layouts list List disk information Display help about menu Display help about the menuing system q Exit from menus Select an operation to perform 1 Add or initialize disks Menu VolumeManager Disk AddDisks Use this operation to add one or more disks to a disk group You can add the selected disks to an existing disk group or to a new disk group that will be created as
11. Disk LUN Select Disk LUN File System Version Current Disk Layout Storage Type Disk LUN Create a datastore on a Fibre Channel iSCSI or local SCSI disk or mount an existing Properties Formatting Network File System Ready to Complete Choose this option if you want to create a Network File System Figure 7 9 Select Storage Type Chapter 7 VMware vSphere considerations 127 b Select the LUN to be used for the datastore from the list in Figure 7 10 and click Next E Disk LUN Name Identifier Path ID LUN Capacity Expandable or VMFS Label v Select Disk LUN File System Version LUN Drive Type CGavent Disk Layout IBM Fibre Channel Disk naa 60050 vmhba2 C0 T0 L1 1 Properties Formatting Ready to Complete Figure 7 10 Select Disk LUN c Select the VMFS version as shown in Figure 7 11 and click Next Tip If the datastore must be shared with ESX hosts under version 5 select VMFS 3 E Disk LUN File System Version Select DiskAUN G VMFS 5 c Disk v Select this option to enable additional capabilities such as 2TB support isk Layout VMFS 5 is not supported by hosts with an ESX version older than 5 0 roperties Formatting C VMFS 3 Ready to Complete Select this option if the datastore will be accessed by legacy hosts Figure 7 11 File System Version d Review the current disk layout in the page shown in Figure 7 12 Ensure that the LUN is
12. Percent of MB Figure 11 2 Skew level of an IBM i benchmark workload Example You are planning the configuration of 2 RAID 5 ranks of 300 GB SSD and 22 RAID 5 ranks of 146 GB 15 K RPM HDD for an IBM i workload The usable capacity on SSD ranks is about 3 4 TB and the capacity on the HDD ranks is about 19 TB therefore about 15 of the overall capacity will reside on SSD Assuming that the active data covers all the capacity and based on the skew level graph in Figure 11 2 about 33 of I O with small transfer sizes will be done to SSD To model performance of the IBM i workload with the described DS8000 configuration and by using Easy Tier to relocate the data you might want to use Disk Magic Since the modelling of Easy Tier is presently not supported for IBM i you can use the following workaround assuming that the workload experiences smaller transfer sizes Split the present peak I O in the ratio 33 67 and model the 33 part on the configuration with 2 RAID 5 ranks of 300 GB SSD and the 67 part on 22 RAID 5 ranks of 146 GB 15 K RPM HDD To get an estimation of the response time in the peak calculate the weighted average of response times of the two parts Hot spot management with Easy Tier Using Easy Tier is an option to relocate hot data to SSD for an IBM i workload For more information about using Easy Tier see the Redpaper publication BM System Storage DS8000 Easy Tier REDP 4667 02 Besides usin
13. To see which Host Bus Adapters HBAs are available use the ioscan fn fc command as shown in Example 10 2 The output shows four HBAs in this server identified as dev fcdN where 0 lt N lt 3 Example 10 2 Identifying HBAs with ioscan fnC fc ioscan fnC fc Class I H W Path Driver S W State H W Type Description fc 0 0 3 1 0 fcd CLAIMED INTERFACE HP AB379 60101 4Gb Dual Port PCI PCI X Fibre Channel Adapter FC Port 1 dev fcd0 fc 1 0 3 1 1 fcd CLAIMED INTERFACE HP AB379 60101 4Gb Dual Port PCI PCI X Fibre Channel Adapter FC Port 2 dev fcdl fc 2 0 7 1 0 fcd CLAIMED INTERFACE HP AB379 60101 4Gb Dual Port PCI PCI X Fibre Channel Adapter FC Port 1 dev fcd2 fc 3 0 7 1 1 fcd CLAIMED INTERFACE HP AB379 60101 4Gb Dual Port PCI PCI X Fibre Channel Adapter FC Port 2 dev fcd3 10 4 Identifying WWPNs of HBAs For zoning purposes you need to know the respective WWPN of the HBAs that are going to be connected to the DS8000 On HP UX 11i use the Fibre Channel mass storage utility fcmsutil to display the WWPNs as shown in Example 10 3 The WWPN is displayed in the middle of the listing Example 10 3 Identifying WWPNS with fcmsutil fcmsutil dev fcd0 Vendor ID is 0x1077 Device ID is 0x2422 PCI Sub system Vendor ID is 0x103C PCI Sub system ID is 0x12D7 PCI Mode PCI X 266 MHz ISP Code version 4 0 90 ISP Chip version 3 Topology PTTOPT FABRIC Link Speed 4Gb Local N Port_id is 0x011d00 Previous
14. With 512 command elements on the virtual SCSI client adapter vscsiX of which 2 are used by the adapter 3 are reserved for each VSCSI LUN for error recovery and the rest are used for I O requests you cannot configure more than 22 DS8000 LUNs as backing devices for a given VSCSI adapter pair when using a queue depth of 20 for the virtual SCSI disk device Example 5 10 Changing the queue depth to 20 for a Virtual SCSI Disk backed by a DS8000 volume chdev 1 hdisk2 a queue_depth 20 P hdisk2 changed On the vscsiX configure the dual VIOS setups for the virtual SCSI adapter path time out feature vscsi_path_to It enables the client adapter to check the health of VIOS servicing a particular adapter and detect if a VIOS is not responding to I O requests In such a case the client will failover to an alternate path if the virtual SCSI adapter path time out is configured This feature is configured using the chdev command as shown in Example 5 11 Example 5 11 Changing the virtual SCSI client adapter path time out value in dual VIOS setups chdev 1 vscsi0 a vscsi_path_to 30 P vscsi0 changed Tip Changes applied with the chdev command and the P option require a reboot to become effective 5 3 Important additional considerations This section describes some additional considerations to be taken when attaching a host running an AIX system to a DS8000 storage system 68 IBM System Storage DS8000 Host Attachment and Interoperabil
15. localhost Celerity FC 8xEN Channel 1 gt NVRAM Save and Commit the changes Repeat these steps for all channels Also make a note of the Port Name as it will be needed for switch zoning and host creation on the DS8000 Tip Write down the Port Name for each Channel that you can see in the ATTO Configuration Tool It will be needed in the configuration of the DS8000 8 4 Using the ATTO Configuration Tool The information contained in this section is based on using the GUI of the Mac OS X Server The user should have access to the server console directly or through a VNC connection to use the ATTO Configuration Tool Chapter 8 Apple considerations 153 8 4 1 Paths Launch the ATTO Configuration Tool by double clicking the ATTO Configuration Tool icon located in the Applications folder The ATTO Configuration Tool consists of three sections Device Listing Lists all installed devices Info Configuration Show details of the device selected in the Device Listing section Status Warnings references are displayed To view the DS8000 storage expand the tree in the Device Listing window on the left until you see your device 2107 and disks LUNs The Paths tab in the multipathing window displays path information as shown in Figure 8 1 The lower half of the window displays information regarding the Target Host Bus Adapter and Statistics ADO ATTO Configuration Tool Device Listing f Paths
16. 248 IBM System Storage DS8000 Host Attachment and Interoperability 12 3 Considerations for z OS This section describes program enhancements that z OS has implemented to support the characteristics of the DS8000 and provides guidelines for the definition of parallel access volumes PAVs 12 3 1 Program enhancements for z OS These relevant Data Facility Storage Management Subsystem DFSMS small program enhancements SPEs were introduced in z OS for support of all DS8000 models gt Scalability support gt Large Volume support LV With LV support it is possible to address a capacity up to 3 964 PB To accommodate the needs of installations that require super large volumes IBM developed a volume named Extended Address Volume EAV recognized by the host as 3390 Model A 3390A Read availability mask support Initial program load enhancements DS8000 device definition Read control unit and device recognition for DS8000 Performance statistics Resource Measurement Facility System Management Facilities Migration considerations Coexistence considerations vvvvvvrvvyy Many of these program enhancements are initially available as authorized program analysis reports APARs and Program Temporary Fix PTF for the current releases of z OS and are later integrated into follow on releases of z OS For this reason see the DS8000 Preventive Service Planner PSP bucket for your current release of z OS Scalability support
17. 81 85 86 82 83 84 87 2107 7701 Y YY YY N SIMPLEX BF BF 1182006 2107 FOO IK SYMBOL DEFINITIONS 220000 RR ek O ONLINE PATH AVAILABLE LOGICALLY OFF PHYSICALLY OFF Data set type dependencies on an EAV For EAV all Virtual Storage Access Method VSAM data set types are eligible to be placed on the extended addressing space or cylinder managed space of an EAV volume running on z OS V1 11 This includes all VSAM data types such as Key sequenced data set Relative record data set Entry sequenced data set Linear data set DB2 IBM IMS IBM CICS and IBM zSeries file system zFS data sets The VSAM data sets placed on an EAV volume can either be storage management subsystem SMS or non SMS managed YYYY Y Starting from z OS V1 12 also following Data set types are now supported sequential basic large partitioned PDS PDSE Catalogs and BDAM data sets in the extended addressing space EAS EAV and data set placement dependencies The EAV volume can be theoretically divided into two parts gt The track managed space from cylinder number 1 to cylinder number 65 520 gt The cylinder managed space from number 65 521 to 1 182 006 Cylinders 258 IBM System Storage DS8000 Host Attachment and Interoperability The EAV track managed area supports all type of data sets The EAV cylinder managed area has restrictions which are explained in Data set type dependencies on an EAV For example
18. Although it is possible to change a volume from protected to unprotected or unprotected to protected use care when using the DS CLI to accomplish this task If the volume is not assigned to any IBM i or is non configured you can change the protection However if volume is configured do not change the protection If you do you must first delete the logical volume which returns the extents used for that volume to the extent pool You can then create a new logical volume with the correct protection after a short period of time depending on the number of extents returned to the extent pool Before deleting the logical volume on the DS8000 you must first remove it from the IBM i configuration assuming it is configured This removal is an IBM i task that is disruptive if the disk is in the system ASP or user ASPs 2 32 because it requires an IPL of IBM i to completely remove the volume from the IBM i configuration This task is the same as removing an internal disk from an IBM i configuration Deleting a logical volume on the DS8000 is similar to physically removing a disk drive from an System i You can remove disks from an IASP with the IASP varied off without performing an IPL on the system 11 6 5 Setting the ports and defining host connections for IBM i 222 When defining a host connection for a DS8000 connection to IBM i natively or with VIOS NPIV specify the hosttype iSeries in the DS8000 GUI or DS CLI command When connecting DS8
19. E Partition Basic NTFS Healthy System 33 90GB 33 83GB 99 z Shared Folders amp D Partition Basic NTFS Healthy 12 00GB 11 93GB 99 Local Users and Groups Performance Logs and Alert A Device Manager A aj Storage Sy Removable Storage Disk Defragmenter 24 Disk Management 4 gt E EJ Services and Applications a ZPDisk 2 beU Basic 12 00 GB 12 00 GB Online Unallocated ZPDisk 3 TE Basic DS8000_EXP 2 12 00 GB 12 00 GB NTFS Online Healthy 4 gt W Unallocated J Primary partition Figure 3 7 Volume size before expansion on Windows 2003 Disk Manager view 28 IBM System Storage DS8000 Host Attachment and Interoperability Figure 3 8 shows the volume size before expansion on the Disk Properties view in Windows 2003 DS8000_EXP 2 Properties ME Sharing Security Shadow Copies Quota General AutoPlay Took Hardware erm Type Local Disk File system NTFS I Used space 67 358 720 bytes 64 2 MB Ji Free space 12 813 393 920 bytes 11 9GB Capacity 12 880 752 640 bytes 11 9GB i Dee Disk Cleanup I Compress drive to save disk space M Allow Indexing Service to index this disk for fast file searching Cancel Figure 3 8 Volume size in Disk Properties view Apply The volume size in Figure 3 7 and Figure 3 8 is 11 99 GB which rounded up is 12 GB To expand this volume on the DS8000 use the chfbvol comma
20. Figure 8 2 Path Status at the Target Level 8 4 3 LUN Base Click a LUN such as LUN 1 shown in Figure 8 1 on page 154 to display the LUN multipathing window The icon to the left of the path shows the status as shown in Figure 8 3 Path Status at LUN Level Path connected to selected LUN is functioning normally Path connected to selected LUN failed If configuration is saved Path to selected LUN is set to Alternate for Read and Write functionality X Path to the selected LUN is DISABLED for Read and Write functionality Figure 8 3 Path status at the LUN Level 8 4 4 Path Actions With the pull down menu Paths you have various options to modify path settings and statistics depending on the device that was selected in the Device Listing section See Figure 8 4 for an example of the Paths menu Setup Save Configuration 3S X Delete Configuration Locate 3L le Reset Statistics Refresh 3R Figure 8 4 Paths options Chapter 8 Apple considerations 155 Setup Select Setup if you need to change Load Balancing Policy or Path behavior Load Balancing can be configured on a per LUN or per Target basis depending on which device is highlighted in the Device Listing pane If a Target is selected then all LUNs controlled by it will be configured the same way Important If a LUN is selected in the Device Listing pane then only that LUN will be affected by configuration changes If a Target is selecte
21. gt O Limit Processing This new feature allows I O optimization for the devices that provide this feature gt exT4 Support Red Hat now supports an unlimited number of subdirectories granular time stamping journeying quota extended attribute support and other small features You can use this new feature for creating the file path on the system during the installation because it is now supported gt Network Block Storage Now Red Hat 6 has support for Fibre Channel over Ethernet for 10 Gb s usage with the same protocols To review all the Red Hat 6 features and how to configure the ones listed before see the Red Hat Storage Administration Manual http docs redhat com docs en US Red_Hat_Enterprise_Linux 6 pdf Storage_Administr ation_Guide Red_Hat_Enterprise_Linux 6 Storage_Administration_Guide en US pdf Working with the DM Multipath configuration The multipath background daemon provides a user interface to print and modify the DM Multipath configuration It can be started as an interactive session with the multipathd k command Within this session a variety of options are available Use the help command to get a list of available options The most significant options are demonstrated in the following examples and in 6 2 10 Understanding non disruptive actions on attached hosts on page 106 The show topology command prints out a detailed view of the current DM Multipath configuration including the state of all availabl
22. m Host Cache Configuration V Allocate space for host cache Use maximum available space 98 GB Custom size J C Ea Li After the configuration is complete the storage information needs to be refreshed manually Use the refresh link button to refresh the storage information a e _ nt Figure 7 49 Host cache configuration Click OK when done Tip Space not used as host cache can still be used to store VMs 7 6 Best practices This section covers best practices for datastores and multipathing 7 6 1 Datastores It is advisable to have a one to one relationship between LUNs and datastores gt One LUN should only contain one datastore gt One datastore should not be spread over more than one LUN There is no general advice possible regarding number and size of datastores The matrix in Table 7 2 can help you decide what is best for the individual setup Table 7 2 Datastore decision matrix Concept Pro Con VMs on one datastore can influence each other Less administrative work gt Better scalability Few large datastores Architectural limits More small datastores Fewer VMs per datastore Better workload isolation Use of RDMs allows for NPIV Administrative overhead Need to spread VMs over multiple datastores IBM System Storage DS8000 Host Attachment and Interoperability 7 6 2 Multipathing To utilize all available paths the PSP need
23. physical disks or LUNs Figure 4 1 Storage and network virtualization with IBM PowerVM and VIOS VIOS Figure 4 1 shows a redundant dual VIOS setup on an IBM Power System providing virtual I O resources to client partitions gt LUN Ais directly accessed by the logical partition using a physical Fibre Channel adapter assigned to the partition gt LUN B is mapped to both VIOS using the physical Fibre Channel adapter in the VIOS partition The LUN is used as a virtual SCSI backing device on both VIOS partitions and provided to the client partition as a generic virtual SCSI device through two paths with two configured virtual SCSI server VSCSI and client adapter pairs On the client partition this VSCSI device is displayed as a generic SCSI disk without any storage subsystem specific properties On VIOS the supported storage system specific multipathing driver must be installed to process the I O to the LUN 48 IBM System Storage DS8000 Host Attachment and Interoperability gt LUN C is accessed through virtual Fibre Channel adapters in the client partition with their own specific WWPNs The LUN is directly mapped to the client partitions WWPNs and is not accessed by the virtual I O server The virtual Fibre Channel adapters are using NPIV and are assigned to a physical Fibre Channel port on the VIOS to physically connect to the SAN The client partition s Fibre Channel I O is only passed to the physical Fibre Channel adapte
24. Mapping vol lun A005 01 Chapter 7 VMware vSphere considerations 131 Increasing the capacity of a datastore If the storage usage of the VMs on a datastore grows it might be necessary to increase its capacity which can be achieved in two ways 1 Add an additional extent LUN to the datastore 2 Increase the size of the LUN on which the datastore resides A datastore can be spread over up to 32 extents However ESXi does not use striping to spread the data across all extents Therefore adding extents to existing datastores can lead to performance issues It causes increased administrative effort because all extents of a datastore need to be handled simultaneously for example during array based snapshots Tip An ESXi host can handle a maximum of 256 LUNs To increase the size of the LUN on which the datastore resides perform the following steps 1 Increase the size of the volume a Using the DS Storage Manager Go to FB Volumes Manage existing volumes Select the volume Action gt Increase capacity Back to FB Volumes Main Page Manage Volumes Select the filtering options to use for displaying volum Filter by LSS X a Add to Volume Groups tems f Properties X Delete amp Migrate gt Figure 7 18 Increase volume size by GUI In the pop up window shown in Figure 7 18 enter the new size and click OK b Using the DS CLI
25. POWER7 TDMF Enterprise Storage Server PowerVM Tivoli FICON POWER WebSphere FlashCopy pSeries XIV GDPS Redbooks z OS HyperSwap Redpaper z VM i5 OS Redbooks logo ZIVSE IBM Resource Link z10 IMS Resource Measurement Facility z9 Informix RMF zEnterprise Lotus Notes Smarter Planet zSeries The following terms are trademarks of other companies Intel Intel logo Intel Inside logo and Intel Centrino logo are trademarks or registered trademarks of Intel Corporation or its subsidiaries in the United States and other countries Linux is a trademark of Linus Torvalds in the United States other countries or both Microsoft Windows and the Windows logo are trademarks of Microsoft Corporation in the United States other countries or both Snapshot and the NetApp logo are trademarks or registered trademarks of NetApp Inc in the U S and other countries Java and all Java based trademarks and logos are trademarks or registered trademarks of Oracle and or its affiliates UNIX is a registered trademark of The Open Group in the United States and other countries Other company product or service names may be trademarks or service marks of others x IBM System Storage DS8000 Host Attachment and Interoperability Preface This IBM Redbooks publication addresses host attachment and interoperability considerations for the IBM System Storage DS8000 s
26. T Services and Applications F Disk 2 Basic 12 00 GB 12 00 GB Online Unallocated amp SDisk 3 Basic DS8000_EXP 2 17 99 GB 12 00 GB NTFS 6 00 GB Online Healthy Unallocated 4 2 BB Unallocated Primary partition Figure 3 9 Expanded volume in Disk Manager Figure 3 9 also shows that Disk3 now has 6 GB of unallocated new capacity To make this capacity available for the file system use the diskpart command Using the diskpart command list the volumes select the pertinent volume check the size of the volume extend the volume and check the size again as shown in Example 3 11 Example 3 11 The diskpart command C Documents and Settings Administrator gt diskpart Microsoft DiskPart version 5 2 3790 3959 Copyright C 1999 2001 Microsoft Corporation On computer X346 TIC 3 DISKPART gt list volume Volume Ltr Label Fs Type Size Status Info Volume 0 Z DS8000_EXP NTFS Partition 12 GB Healthy Volume 1 E NTFS Partition 34 GB Healthy System Volume 2 D DVD ROM 0 B Healthy Volume 3 C NTFS Partition 68 GB Healthy Boot DISKPART gt select volume 0 Volume 0 is the selected volume DISKPART gt detail volume 30 IBM System Storage DS8000 Host Attachment and Interoperability Disk Status Size Free Dyn Gpt Disk 3 Online 18 GB 6142 MB Readonly No Hidden No No Default Drive Letter No Shadow Copy No DISKPART gt extend DiskPart
27. Use the DS CLI command chfbvol as in Example 7 3 and specify the new capacity and the volume ID Example 7 3 increase capacity using chfbvol dscli gt chfbvol cap 40 A000 Date Time 31 Mai 2012 13 33 37 CEST IBM DSCLI Version 7 6 20 221 DS IBM 2107 75TV181 CMUC0O0332W chfbvol Some host operating systems do not support changing the volume size Are you sure that you want to resize the volume y n y CMUC00026I chfbvol FB volume A000 successfully modified 132 IBM System Storage DS8000 Host Attachment and Interoperability Increase the size of the VMFS datastore in vSphere a Go to Configuration Storage and click Rescan All b Next select the datastore Right click and select Properties See Figure 7 19 for an example View Datastores Devices Datastores Refresh Delete Add Storage Rescan All Browse Datastore Alarm gt Assign User Defined Storage Capability Rename Location Unmount Hardware 920 00MB W Used Delete 18 85GB E Free TENY Refresh z System S Open in New Window Ctri Alt N User defin Refresh Path oi Round Figure 7 19 Datastore properties Fibre Channel Disk naa 20 00 GB Copy to Clipboard c In the volume properties window shown in Figure 7 20 click Increase Datastore Name DS8000_Datastorel File System VMFS 5 54 Total Capacity 19 75 GB Maximum File Size 2 00 TB Block Size 1M
28. Use this mode if the LUN also needs to be used by a physical server such as for a cluster Virtual The RDM can be included in snapshots and can benefit from other virtual disk features The compatibility mode you choose will apply only to this virtual disk and will not affect any other disks using this LUN mapping Compatibility Physical Allow the guest operating system to access the hardware directly Taking a snapshot of this virtual machine will not indude this disk C Virtual Allow the virtual machine to use VMware snapshots and other advanced functionality Figure 7 35 Compatibility mode for RDM 8 In the advanced options in Figure 7 36 the virtual device node can be specified It specifies where in the virtual SCSI bus of the VM the RDM should be attached Specify the advanced options for this virtual disk These options do not normally need to be changed Virtual Device Node o scst 0 1 X rs fie 0 0 Figure 7 36 virtual device node for RDM 9 Verify all options in the Ready to Complete window and click Finish Options Hardware type Hard Disk Create disk Use mapped system LUN Virtual Device Node SCSI 0 1 Disk mode Persistent Target LUN IBM Fibre Channel Disk naa 600507630affc29f000000000000a8005 Compatibility mode Physical Mapped datastore Store with VM Figure 7 37 Ready to Complete window for RDM 10 Close the Virtual Machine Properties wi
29. as listed in 11 1 4 Useful websites on page 204 212 IBM System Storage DS8000 Host Attachment and Interoperability 11 3 6 Number of Fibre Channel adapters in IBM i and in VIOS When deciding on the number of Fibre Channel adapters in an IBM i and DS8000 configuration in addition to the maximal number of LUNs per port specified in Table 11 3 on page 207 also consider the performance aspect of the configuration To achieve good performance of DS8000 with IBM i in some cases you will need to implement less than the maximum number of LUNs per port Here the most important factor to consider is the adapter throughput capacity This section provides the measured maximal I O rate in IOPS and maximal data rate MB per second or MBps per an IOP based adapter or per port in an IOP less adapter Based on these values information is provided about the sizing guidelines for the capacity number of LUNs per port or per adapter to achieve good performance Also provided are guidelines for sizing Fibre Channel adapters in IBM i for native attachment to the DS8000 and for sizing Fibre Channel adapters in VIOS for VIOS VSCSI or VIOS NPIV attachment Fibre Channel adapters in native attachment For the measured values of maximal I O rates and data rates per port in other adapters see the following rates gt A4GBIOP based Fibre Channel adapter feature number 5760 connected to the IOP feature number 2844 can support these maximums
30. if you allocate a 3390 Model 9 with 65 520 cylinders and place sequential data sets on it Figure 12 3 perform a dynamic volume expansion because of a volume full condition to make the volume into an EAV as 3390 Model A The required VTOC reformat is successfully performed First 65 520 cylinders VTOC VTOC Index VVDS ICF Catalog Page Data Set SMS and nonVSAM NonSMS VSAM Keyrange Cyl 65521 Imbed haan Sinks Figure 12 3 Data set placement on EAV gt Breakpoint lt Breakpoint VSAM preferred For software trying to allocate an extent of sequential data sets will fail and produce the error IEF2571 Space requested not available even if there is a surplus of space physically available in the cylinder managed area Data sets with extents on the cylinder managed area are not supported with z OS V1 10 Placing any kind of VSAM data set on the same volume will not cause allocations errors 12 4 2 z OS prerequisites for EAV EAV volumes have the following prerequisites gt EAV volumes are only supported on z OS V1 10 and later If you try to bring an EAV volume online for a system with a release older then z OS V1 10 the EAV volume does not come online gt The 3390 Model A definitions have no additional HCD considerations gt On parameter library parmlib member IGDSMSxx the parameter USEEAV YES must be set to allow data set allocations on EAV volumes The defa
31. read write available 255 0 0 100 3 3 1016 6 16 2301 0 2301 0 0 0 dev disk disk12 available 767 767 On dev disk disk13 available 767 767 On dev disk disk14 available 767 767 On Chapter 10 HP UX considerations 189 The vgcreate command takes many arguments which can be quite confusing A brief breakdown of the options used in these examples is as follows y Setting this options to y will automatically back up the changes made to the volume group This causes the vgcfgbackup command to be executed after the vgcreate command X y Setting this options to y allows the allocation of more physical extents on the physical volume 1 255 This argument determines the maximum number of logical volumes that the volume group can contain The allowable range is 1 to 255 the default is 255 p 100 This argument determines the maximum number of physical volumes that the volume group can contain The allowable range is 1 to 255 the default is 16 s 16 This argument determines the number of megabytes in each physical extent The allowable range is 1 to 256 expressed in powers of 2 2 4 8 16 The default is 4MB Caution Do not use the f option on the vgcreate command It forces the creation of a volume group with a physical volume that has alternate blocks already allocated It can be a dangerous command because the potential for data corruption exists 10 6 2 Exposing link e
32. using the DSCLI mkhostconnet wwname 10000000c9831fa5 hosttype iseries volgrp vl i_portl 11 6 6 Zoning the switches When zoning the switches that connect the DS8000 natively to IBM i consider the following guidelines gt Zone the switches so that each port in IBM i Fibre Channel adapters accesses one port in DS8000 Multiple ports from IBM i adapters can access one port in DS8000 You can share a DS8000 with I O traffic from multiple IBM i systems or other servers providing that the aggregated 1 O traffic from attached systems do not overload the port gt Do not zone the switches so that multiple IBM i ports can access multiple DS8000 ports because IBM i establishes only one path from one port to the disk If you do all IBM i ports establish the path through the same DS8000 port but leave other DS8000 ports unused causing performance bottlenecks on the used port 11 6 7 Setting up VIOS As opposed to connecting with VIOS VSCSI with VIOS NPIV you do not need to map the LUNs in VIOS to the ports in virtual Fibre Channel adapters Instead you assign the LUNs to virtual Fibre Channel ports when defining host connections in the DS8000 Chapter 11 IBM i considerations 223 VIOS VSCSI To implement multipath with two VIOS complete the following steps 1 Remove the SCSI reservation attribute from the LUNs or hdisks that are to be connected through two VIOS To do this enter the following command for each hdisk that will co
33. 0000 pci 0000 pci 0000 pci 0000 le le 1c le lec le 1c le 24 24 24 24 24 24 24 24 1s 1 dev disk by path cut c 44 0 fc 0x500507630a00029f 0 fc 0x500507630a00029f 0 fc 0x500507630a00029f 0 fc 0x500507630a00029f 0 fc 0x500507630a0b029f 0 fc 0x500507630a0b029f 0 fc 0x500507630a0b029f 0 fc 0x500507630a0b029f 0 fc 0x500507630a13029f 0 fc 0x500507630a13029f 0 fc 0x500507630a13029f 0 fc 0x500507630a13029f 0 fc 0x500507630a1b029f 0 fc 0x500507630a1b029f 0 fc 0x500507630a1b029f 0 fc 0x500507630a1b029f 0x0000000000000000 0x0001000000000000 0x0002000000000000 0x0003000000000000 0x0000000000000000 0x0001000000000000 0x0002000000000000 0x0003000000000000 0x0000000000000000 0x0001000000000000 0x0002000000000000 0x0003000000000000 0x0000000000000000 0x0001000000000000 0x0002000000000000 0x0003000000000000 Chapter 6 Linux considerations 6 o o o o o o oe e gt o e ASNT Sa SS OTS SS Da o SS OS sdb sdc sdd sde sdf sdg sdh sdi sdj sdk sdl sdm sdn sdo sdp sdq 95 6 2 7 Adding DS8000 volumes to Linux on System z Only in the most recent Linux distributions for System z does the zfcp driver automatically scan for connected volumes This section provides information about how to configure the system so that the driver automatically
34. 11 in dev mapper x36501ab9 1s 1 dev mapper cut c 48 3600507630af fc29f0000000000004690 3600507630af fc29f0000000000004691 3600507630af fc29F0000000000004790 3600507630af fc29f0000000000004791 Important The device mapper itself creates its default device nodes in the dev directory These nodes are named dev dm 0 dev dm 1 and so on However these nodes are not persistent can change with configuration changes and cannot be used for device access SLES 11 creates an additional set of device nodes for multipath devices It overlays the former single path device nodes in the dev disk by id directory It means that any device mapping for example mounting a file system configured with one of these nodes works exactly the same as before starting DM Multipath Instead of using the SCSI disk device to start it uses the DM Multipath device as shown in Example 6 30 Example 6 30 SLES 11 DM Multipath device nodes in dev disk by id x36501ab9 1s 1 dev disk by id cut c 44 scsi 3600507630af fc29f0000000000004690 gt dm 0 scsi 3600507630af fc29f0000000000004691 gt dm 3 scsi 3600507630af fc29f0000000000004790 gt dm 1 scsi 3600507630af fc29f0000000000004791 gt dm 2 If you set the user_friendly_names option in the etc multipath conf file SLES 11 will create DM Multipath devices with the names mpatha mpathb and so on in the dev mapper directory The DM Multipath device nodes in the
35. 115 Disk dev sdb doesn t contain a valid partition table Disk dev sdd 53 7 GB 53687091200 bytes 64 heads 32 sectors track 51200 cylinders Units cylinders of 2048 512 1048576 bytes Disk identifier 0x00000000 Disk dev sdd doesn t contain a valid partition table 6 5 2 Monitoring performance with the iostat command Use the iostat command to monitor the performance of all attached disks It is part of the sysstat package that ships with every major Linux distribution but is not necessarily installed by default The iostat command reads data provided by the kernel in proc stats and prints it in readable format For more details see the man page for iostat 6 5 3 Working with generic SCSI tools For Linux there is a set of tools that allow low level access to SCSI devices These tools are called the sg tools They communicate with SCSI devices through the generic SCSI layer which is represented by the dev sg0 dev sg1 and so on special device files In recent Linux versions the sg_tools can also access the block devices dev sda dev sdb or any other device node that represents a SCSI device directly The following sg_tools are useful gt sg_ing dev sgx prints SCSI inquiry data such as the volume serial number gt sg_scan prints the SCSI host channel target and LUN mapping for all SCSI devices gt sg_map prints the dev sdx to dev sgy mapping for all SCSI devices gt sg _readcap dev sgx prints the blo
36. 1284 0 3 fscsil hdisk14 OPEN NORMAL 1347 0 DEV 1 DEVICE NAME vpathl TYPE 2107900 POLICY Optimized SERIAL 75TV1812201 Path Adapter Hard Disk State Mode Select Errors 0 fscsi0 hdisk3 OPEN NORMAL 927 0 1 fscsi0 hdisk7 OPEN NORMAL 985 0 2 fscsil hdisk11 OPEN NORMAL 940 0 3 fscsil hdisk15 OPEN NORMAL 942 0 Here are some additional features of SDD for AIX gt Enhanced SDD configuration methods and migration SDD has a feature in the configuration method to read the PVID from the physical disks and convert the PVID from hdisks to vpaths during the SDD vpath configuration With this feature you can skip the process of converting the PVID from hdisks to vpaths after configuring SDD devices Furthermore SDD migration can skip the PVID conversion process This tremendously reduces the SDD migration time especially with a large number of SDD devices and in an logical volume manager LVM configuration environment Migration option for large device configuration SDD offers an environment variable SKIP_SDD_MIGRATION to customize the SDD migration or upgrade to maximize performance The SKIP_SDD_ MIGRATION environment variable is an option available to permit the bypass of the SDD automated migration process backup restoration and recovery of LVM configurations and SDD device configurations This variable can help decrease the SDD upgrade time if you reboot the system after upgrading SDD 5 4 2 SDDPCM for AIX The base functionali
37. 2 root prime lab 04 df devices system contract ctfs dev dsk c0t0d0s0 21248478 blocks 1706174 files devices 0 blocks 0 files 0 blocks 2147483598 files ye proc proc 0 blocks 29889 files etc mnttab mnttab 0 blocks 0 files etc svc volatile swap 22660000 blocks 567860 files system object objfs 0 blocks 2147483414 files dev fd fd 0 blocks 0 files var dev dsk cOt0d0s5 2064656 blocks 591399 files tmp swap 22660000 blocks 567860 files var run swap 22660000 blocks 567860 files export home dev dsk cO0t0d0s7 53769918 blocks 4725414 files 172 IBM System Storage DS8000 Host Attachment and Interoperability For these examples a Fujitsu Siemens Computers Primepower 450 server equipped with two LP10000 Emulex HBAs was used as you can see in Example 9 21 Taking it into account you can find these processes in the Emulex Bootcode User Manual Boot Code User Manual For Emulex Adapters at this website http www emulex com files downloads hardware bootcode pdf Follow the steps described in the topic Configure Boot from SAN on Solaris SFS SPARC of the Emulex document All the steps need to be executed from the Open Boot Prompt OBP environment thus without running Solaris This document instructs that boot code must be installed and enabled to boot from SAN Also note that the boot code for Solaris emlxs SFS systems in enabled automatically when it is installed
38. 2 fp 0 0 ssd w500507630a03029F 0 3 3t500507630A03029Fd1 lt IBM 2107900 268 cyl 8736 alt 2 hd 30 sec 64 gt pci 80 2000 em1 x 2 fp 0 0 ssd w500507630a03029F 1 4 c3t500507630A03029Fd2 lt IBM 2107900 268 cyl 17918 alt 2 hd 64 sec 256 gt pci 80 2000 em1 x 2 fp 0 0 ssd w500507630a03029F 2 5 c4t500507630A00029Fd0 lt IBM 2107900 268 cyl 4367 alt 2 hd 30 sec 64 gt pci 80 2000 em1 x 1 fp 0 0 ssd w500507630a00029Ff 0 6 c4t500507630A00029Fd1 lt IBM 2107900 268 cyl 8736 alt 2 hd 30 sec 64 gt pci 80 2000 em1x 1 fp 0 0 ssd w500507630a00029F 1 7 c4t500507630A00029Fd2 lt IBM 2107900 268 cyl 17918 alt 2 hd 64 sec 256 gt pci 80 2000 em1x 1 fp 0 0 ssd w500507630a00029F 2 3 Customize the kernel drv scsi_vhci conf file as shown in Example 9 26 Example 9 26 Customizing the kernel drv scsi_vhci conf file bash 3 00 more kernel drv scsi_vhci conf Copyright 2004 Sun Microsystems Inc All rights reserved Use is subject to license terms pragma ident scsi_vhci conf 1 9 04 08 26 SMI name scsi_vhci class root Load balancing global configuration setting load balance none will cause all I O to a given device which supports multipath I 0 to occur via one path Setting load balance round robin will cause each path to the device to be used in turn load balance round robin Automatic failback configuration possible values are auto failback enable or auto failba
39. 2 hd 30 sec 64 gt scsi_vhci ssd g600507630af fc29f0000000000003101 4 c5t600507630AFFC29F0000000000003102d0 lt IBM 2107900 268 cyl 17918 alt 2 hd 64 sec 256 gt scsi_vhci ssd g600507630af fc29f0000000000003102 Specify disk enter its number 9 6 4 Associating the MPxIO device file and underlying paths Example 9 29 shows the association between the MPxIO device special file representation and its underlying paths Example 9 29 MPxIO device file representation and underlying paths bash 3 00 luxadm display dev rdsk c5t600507630AFFC29F0000000000003101d0s2 DEVICE PROPERTIES for disk dev rdsk c5t600507630AFFC29F0000000000003101d0s2 Vendor IBM Product ID 2107900 Revision 268 Serial Num 751TV1813101 Unformatted capacity 8192 000 MBytes Write Cache Enabled Read Cache Enabled Minimum prefetch 0x0 Maximum prefetch 0x16 Device Type Disk device Path s dev rdsk c5t600507630AFFC29F0000000000003101d0s2 devices scsi_vhci ssd g600507630af fc29f0000000000003101 c raw Controller devices pci 80 2000 em1 x 1 fp 0 0 Device Address 500507630a00029f 1 Host controller port WWN 10000000c95f3592 Class primary State ONLINE Controller devices pci 80 2000 em1 x 2 fp 0 0 Device Address 500507630a03029f 1 Host controller port WWN 10000000c95f35dc Class primary State ONLINE Chapter 9 Solaris considerations 179 180 IBM System Storage DS8000 Host Attachment and Interoperability 10 HP UX considerations This
40. 21 Example 6 21 Specify user friendly names in etc multipath conf defaults user_friendly_names yes The names generated using this method are persistent meaning that they do not change even if the device configuration changes If a volume is removed its former DM Multipath name will not be used again for a new one If it is re attached it will get its old name The mappings between unique device identifiers and DM Multipath user friendly names are stored in the file var 1ib multipath bindings User friendly names The user friendly names are separate for SLES 11 and Red Hat Enterprise Linux 5 No new changes were found for the user friendly set up between Red Hat Enterprise Linux 5 and 6 however there are some minor difference for enabling multipathing For more information about working with user friendly names see Accessing DM Multipath devices in SLES 11 on page 102 and Accessing DM Multipath devices in Red Hat Enterprise Linux 5 on page 103 Chapter 6 Linux considerations 99 Enabling multipathing for SLES 11 You can start device mapper multipathing by running two start scripts that have already been prepared as shown in Example 6 22 Example 6 22 Starting DM Multipath in SLES 11 x36501ab9 etc init d boot multipath start Creating multipath target done x36501ab9 etc init d multipathd start Starting multipathd done In order to have DM Multipath start automatically at each system start add the s
41. 24 ff 2d c5 68 21 00 00 24 ff 2d c5 68 vmhbas Fibre Channel 20 00 00 24 ff 2d c5 69 21 00 00 24 ff 2d c5 69 Software LsI1078 Licensed Features F wmbhat Riar EOST Time Configuration Details DNS aa Raig l hba2 Authentication Services Model ISP2532 based 8Gb Fibre Channel to PCI Express HBA Power Management WWN 20 00 00 24 ff 2d c3 40 21 00 00 24 ff 2d c3 40 Virtual Machine Startup Shutdown Targets 2 Devices 2 Paths 4 Virtual Machine Swapfile Location View Devices Paths Security Profile a LUN Type Host Cache Configuration Runtime Name 3 507630affc29f0000000 System Resource Allocation 9302 a 600507630affc29f0000000 Agent VM Settings Advanced Settings Copy identifier to clipboard Figure 7 43 Show LUNs attached to ESXi host 2 From the drop down list at the top select Round Robin VMware and click Change Figure 7 44 IBM Fibre Channel Disk naa Policy Path Selection Most Recently Used VMware h Most Recently Used VMware Storage Array Type Round Robin VMware Fixed Eeo S O O Paths Runtime Name Target yvmhba5 C0 T0 L0 50 05 07 63 0a ff c2 9f 50 05 07 63 0a 00 42 9F 3 vmhba5 C0 T1 L0 50 05 07 63 0a ff c2 9f 50 05 07 63 0a 40 42 9f 0 vmhba4 C0 T1 L0 50 05 07 63 0a ff c2 9f 50 05 07 63 0a 40 42 9F 0 Active a vmhba4 C0 T0 L0 50 05 07 63 0a ff c2 9f 50 05 07
42. 3rd party symmetric device need an entry similar to following in this file Just replace the SUN SENA part with the Vendor ID Product ID for the device exactly as reported by Inquiry cmd device type scsi options list SUN SENA symmetric option symmetric option 0x1000000 Chapter 9 Solaris considerations 175 To supply the VID PID string complete the following steps 1 Disable MPxIO as shown in Example 9 24 Example 9 24 Disabling MPxIO bash 3 00 stmsboot d WARNING stmsboot operates on each supported multipath capable controller detected in a host In your system these controllers are devices pci 80 2000 em1 x 2 fpe0 0 devices pci 80 2000 em1x 2 fp 1 0 devices pci 80 2000 em1x 1 fpe0 0 devices pci 80 2000 em1x 1 fp 1 0 If you do NOT wish to operate on these controllers please quit stmsboot and re invoke with D fp mpt to specify which controllers you wish to modify your multipathing configuration for Do you wish to continue y n default y Checking mpxio status for driver fp Checking mpxio status for driver mpt WARNING This operation will require a reboot Do you want to continue y n default y The changes will come into effect after rebooting the system Reboot the system now y n default y updating platform sun4us boot_archive 2 Restart your system Example 9 25 shows that after restarting your system the following actions were taken The fi
43. 6000 4000 2000 A Mm wok ang sw sui asuodsas abelaay Total reads thousands Oo Figure 11 3 Read operations on IBM i objects by iDoctor gt IBM i Media Preference Data Migration is the method to migrate specific IBM i objects to SSD by using IBM i commands such as Change Physical File CHGPF with parameter UNIT SSD gt ASP balancing is the IBM i method based on the data movement within a system disk pool or base disk pool Auxiliary Storage Pool ASP in IBM i Two hot spot management methods are available with ASP balancing Hierarchical Storage Management HSM balancing In this way of data relocation you first trace the ASP by command TRCASPBAL which collects the data statistics then you use the command STRASPBAL with parameter TYPE HSM to actually relocate the hot data to high performing LUNs Media Preference MP balancing This ASP balancer function helps correct any issues with media preference flagged database objects which are not on their preferred media type which is either SSDs or HDDs You start it with the command STRASPBAL TYPE MP with the parameter SUBTYPE set to CALC SSD or HDD depending which way of data migration that you want to use MP balancing is supported starting with IBM i V7 1 For more information about using IBM i based tools and methods see the Redbooks publication DS8800 Performance Monitoring and Tuning SG24 8013 00 and the IBM i Information Center
44. 64 LPARs 255 unit checks to each of 64 LPARs 4 193 820 stage changes total 16 384 unit checks total Host state change processing Host unit check processing Receive unsolicited 85 status Receive unsolicited unit check Spawn AOM task Issue Sense Issue Sense ID RDC RCD SNSS Spawn AOM task Issue Read Feature Codes Issue PSF RSSD for Query 55 Issue Query PPRC State if cascaded Issue Reset Summary Unit Check Figure 12 5 PPRC error event comparison IBM System Storage DS8000 Host Attachment and Interoperability 12 6 z VM considerations This section provides specific considerations that are relevant when attaching a DS8000 series to a z VM environment 12 6 1 Connectivity The z VM operating system provides the following connectivity gt z VM supports FICON attachment as 3990 Model 3 or 6 controller gt Native controller modes 2105 and 2107 are supported on z VM V5 4 The z VM operating system also simulates controller mode support by each guest gt z VM supports FCP attachment for Linux systems running as a guest gt z VM supports FCP attached SCSI disks starting with z VM V5 4 and emulated FBA devices 12 6 2 Supported DASD types and LUNs The z VM operating system supports the following extended count key data IBM ECKD DASD types gt 3390 Models 2 3 and 9 including the 32 760 and 65 520 cylinder custom volumes gt 3390 Model 2 and 3 in 3380 track compatibility mode gt 3390 Mod
45. Basic Disks Dynamic Disks Other Disk Subsystem uw HDDs HB Hardware E Microsoft Functionality C Non Microsoft Functionality DS8000 DS6000 Figure 3 11 Example of Microsoft VDS architecture For a detailed description of VDS see the Microsoft Windows Server 2003 Virtual Disk Service Technical Reference found at the following website http technet microsoft com en us ibrary cc776012 WS 10 aspx The DS8000 can act as a VDS hardware provider The implementation is based on the DS8000 common information model CIM agent a middleware application that provides a CIM compliant interface The VDS uses CIM technology to list information and manage LUNs For information about how to install and configure VDS support see the BM System Storage DS Open Application Programming Interface Reference GC35 0516 3 9 1 VDS integration with DS8000 storage subsystems The following sections present examples of VDS integration with advanced functions of the DS8000 storage subsystems when leveraging the DS CIM agent 3 9 2 Volume Shadow Copy Service The Volume Shadow Copy Service also known as Volume Snapshot Service VSS provides a mechanism for creating consistent real time copies of data known as shadow copies It integrates IBM System Storage FlashCopy to produce consistent shadow copies while also coordinating with business applications file system services backup applications and fast recovery solutions For mor
46. CD ROM ANSI SCSI revision 05 Host scsi3 Channel 00 Id 00 Lun 00 Vendor IBM Model 2107900 Rev 288 Type Direct Access ANSI SCSI revision 05 Host scsi3 Channel 00 Id 00 Lun 01 Vendor IBM Model 2107900 Rev 288 Type Direct Access ANSI SCSI revision 05 Host scsi3 Channel 00 Id 00 Lun 02 Vendor IBM Model 2107900 Rev 288 Type Direct Access ANSI SCSI revision 05 Host scsi3 Channel 00 Id 00 Lun 03 Vendor IBM Model 2107900 Rev 288 Type Direct Access ANSI SCSI revision 05 Host scsi3 Channel 00 Id 01 Lun 00 Vendor IBM Model 2107900 Rev 288 Type Direct Access ANSI SCSI revision 05 The fdisk 1 command can be used to list all block devices including their partition information and capacity but it does not contain the SCSI address vendor and model information To view the list see Example 6 59 Example 6 59 Output of the fdisk I command x36501ab9 fdisk 1 Disk dev sda 146 7 GB 146695782400 bytes 255 heads 63 sectors track 17834 cylinders Units cylinders of 16065 512 8225280 bytes Disk identifier 0x5bbe5807 Device Boot Start End Blocks Id System dev sdal 1 5221 41937651 83 Linux dev sda2 5222 10442 41937682 83 Linux dev sda3 10443 11486 8385930 82 Linux swap Solaris Disk dev sdb 53 7 GB 53687091200 bytes 64 heads 32 sectors track 51200 cylinders Units cylinders of 2048 512 1048576 bytes Disk identifier 0x00000000 Chapter 6 Linux considerations
47. Control amp Local Users and Groups Storage YB Windows Server Backup g Disk Management 4 Disk 4 Dynamic New Volume I 10 00 GB 10 00 GB NTFS Online Healthy Disk 5 Dynamic New Volume I 9 97 GB 9 97 GB NTFS Online Healthy LaDisk 6 Dynamic New Volume K 9 97 GB 9 97 GB NTFS Online Healthy Disk 7 SSS Dynamic New Volume K 9 97 GB 9 97 GB NTFS Online Healthy Figure 3 6 Server Manager view 26 IBM System Storage DS8000 Host Attachment and Interoperability 3 7 Expanding dynamic disk for Windows 2003 Windows 2008 It is possible to expand a volume in the DS8000 even if it is mapped to a host Operating systems such as Windows 2003 and Windows 2008 can handle the volumes being expanded even if the host has applications running A volume that is in an IBM FlashCopy Metro Mirror or Global Mirror relationship cannot be expanded unless the relationship is removed Before you expand the volume remove FlashCopy Metro Mirror or Global Mirror on that volume If the volume is part of a Microsoft Cluster Service MSCS Microsoft advises that you shut down all nodes except one Applications in the resource using the volume that is going to be expanded must be stopped before expanding the volume However applications running in other resources can continue After expanding the volume restart the applications and the resource and then restart
48. Corporation North Castle Drive Armonk NY 10504 1785 U S A The following paragraph does not apply to the United Kingdom or any other country where such provisions are inconsistent with local law INTERNATIONAL BUSINESS MACHINES CORPORATION PROVIDES THIS PUBLICATION AS IS WITHOUT WARRANTY OF ANY KIND EITHER EXPRESS OR IMPLIED INCLUDING BUT NOT LIMITED TO THE IMPLIED WARRANTIES OF NON INFRINGEMENT MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE Some states do not allow disclaimer of express or implied warranties in certain transactions therefore this statement may not apply to you This information could include technical inaccuracies or typographical errors Changes are periodically made to the information herein these changes will be incorporated in new editions of the publication IBM may make improvements and or changes in the product s and or the program s described in this publication at any time without notice Any references in this information to non IBM websites are provided for convenience only and do not in any manner serve as an endorsement of those websites The materials at those websites are not part of the materials for this IBM product and use of those websites is at your own risk IBM may use or distribute any of the information you supply in any way it believes appropriate without incurring any obligation to you Any performance data contained herein was determined in a controlled environment Therefore the res
49. DS8000 volume follow these steps 1 When the QLogic HBAs are initialized during the server startup process press Ctrl Q Figure 2 2 When the message lt CTRL Q gt Detected Initialization in progress Please wait is displayed the QLogic BIOS setup utility is being loaded QLogic Corporation QLE2468 PCI Fibre Channel ROM BIOS Version 2 82 Copyright C QLogic Corporation 1993 2088 ALI rights reserved wuu gqlogic com Press lt CTRL gt or lt ALT Q gt for Fast UTIL lt CTRL gt Detected Initialization in progress Please wait BIOS for Adapter is disabled BIOS for Adapter 1 is disabled ROM BIOS NOT INSTALLED Figure 2 2 Entering the QLogic BIOS setup utility Use these methods to navigate the menu items Press the arrow keys to move between menu items Press Enter to select or change an item Press Esc to leave a window and return to the previous level 2 When the QLogic BIOS setup utility opens select a host adapter if more than one is installed If there is only one installed it will be selected by default After the HBA that you want to configure is selected press Enter 3 In the main Fast UTIL Options panel upper panel of Figure 2 3 leave the cursor on Configuration Settings default and press Enter 4 In the Configuration Settings panel middle panel of Figure 2 3 leave the cursor on Adapter Settings default and press Enter Chapter 2 Open systems cons
50. Device 1 0 Adapter 0 ooo F12 Cancel 1 0 Bus 0 0 0 0 Ctl Dev 0 3 0 2 0 1 0 0 Figure 11 38 Selecting the load source device 3 After confirming the selection of the load source disk select option 2 Install Licensed Internal Code and Initialize system Figure 11 39 to proceed with IBM i installation Select one of the following aoFWwWNr re Selection 2 Serial Number Type 50 210A29F 2107 F3 Exit F12 Cancel Model A04 Restore Licensed Internal Code Install Licensed Internal Code and Initialize system Install Licensed Internal Code and Recover Configuration Install Licensed Internal Code and Restore Disk Unit Data Install Licensed Internal Code and Upgrade Load Source I 0 Bus Install Licensed Internal Code LIC Disk selected to write the Licensed Internal Code to Controller 0 Device 3 Figure 11 39 Installing Licensed Internal Code and initialize system 11 8 2 Booting from SAN and cloning When booting from SAN support you can take advantage of certain advanced features that are available with the DS8000 series and copy services functions These functions provide the ability to perform a rapid copy of the data held on DS8000 LUNs Therefore when you have a system that only has external LUNs with no internal drives you can create a clone of your system Important For the instances in this section a clone is a copy of a system that only uses external
51. Disk Management L Disk 0 Basic System Reserved 556 93 GB 100 MB NTFS Online Healthy System Active Primary Partition C 556 83 GB NTFS Healthy Boot Page File Crash Disk 3 Unknown 20 00 GB 20 00 GB Offline i Unallocated Help Figure 3 23 Disk management IBM System Storage DS8000 Host Attachment and Interoperability 2 In Hyper V manager select the virtual machine right click and select Settings Hyper Manager es Almlal J File Action view Window Help 34 Hyper V Manager aa WIN GD307VJHE1L irtual Machines Name State Running Shut Down Save Figure 3 24 Virtual machine settings 3 Select the controller to which the new VHD needs to be attached select Hard Drive and click Add B Settings for Windows 2003 amp Hardware 1 Add Hardware JE B105 Boot from CD m Memory 1024 MB B Processor 1 Virtual processor EE IDE Controller 0 Hard Drive Windows 2003 vhd E IDE Controller 1 DVD Drive est iso E SCSI Controller U Network Adapter windows 2003 7 4 Q BE SCSI Controller You can add hard drives to your SCSI controller or remove the SCSI controller from the machine Click on the Add button to add a new hard drive to this SCSI controller To remove the virtual storage adapter from this virtual machine click Remove Controller All virtual hard
52. E anaa 160 Device Specific Z0 10 Device Specific Z1 0135 Device Specific Z2 075 Device Specific Z3 22010 Device Specific Z4 08 Device Specific Z5 00 PLATFORM SPECIFIC Name disk Node disk Device Type block Ispath dev hdisk0 status name parent connection Enabled hdiskO fscsi0 500507630a4b429f 4090400000000000 Enabled hdiskO fscsi0 500507630a40429f 4090400000000000 Enabled hdiskO fscsil 500507630a4b429f 4090400000000000 Enabled hdiskO fscsil 500507630a40429f 4090400000000000 The Ispath command provides information about the available paths for a given LUN when using AIX MPIO with SDDPCM It displays the Fibre Channel adapter port in the server fscsi0 the DS8000 I O port WWPN 500507630a4b429f and the DS8000 volume ID 9000 with third fourth sixth and eight positions in the number 4090400000000000 provided with the output for each path 54 IBM System Storage DS8000 Host Attachment and Interoperability By using the DSCLI 1sioport command as shown in Example 4 7 you can identify the I O ports used on the DS8000 for each path Example 4 7 Using the DSCLI Isioport command to list the WWPNs of the DS8000 I O ports dscli gt lsioport Date Time 30 de mayo de 2012 10 39 40 AM CEST IBM DSCLI Versio IBM 2107 75TV181 ID WWPN State Type topo portgrp 10005 500507630A40429F Online Fibre Channel SW SCSI FCP 0 10135 500507630A4B429F Online Fibre Channel SW SCS
53. Example 6 44 shows how to add two DS8000 ports to the second HBA Example 6 44 Attaching remote ports to the second HBA Inxmnt01 echo 0x500507630a08029f gt sys bus ccw devices 0 0 0600 port_add Inxmnt01 echo 0x500507630a03029f gt sys bus ccw devices 0 0 0600 port_add Inxmnt01 Is sys bus ccw devices 0 0 0600 grep Ox 0x500507630a03029F 0x500507630a08029F 6 3 Resizing DS8000 volumes dynamically Currently only SLES11 SP1 and RHEL 6 are capable of utilizing the additional capacity of dynamically enlarged DS8000 volumes Reducing the size is not supported for versions before RHEL 5 To resize DS8000 volumes dynamically complete the following steps 1 Create an ext3 file system on one of the DS8000 multipath devices and mount it The df command demonstrated in Example 6 45 shows the available capacity Example 6 45 Checking the size and available space on a mounted file system x36501ab9 df h mnt itso_4791 Filesystem Size Used Avail Use Mounted on dev mapper 3600507630af fc29f0000000000004791 50G 180M 47G 1 mnt itso_4791 2 Use the chfbvol command of the DSCLI to increase the capacity of the volume from 50 to 100 GB as shown in Example 6 46 Chapter 6 Linux considerations 109 110 Example 6 46 Increasing the size of a DS8000 volume dscli gt chfbvol cap 100 4791 Date Time October 12 2010 4 48 59 PM CEST IBM DSCLI Version CMUC00332W chfbvol Some host operating systems do not support ch
54. Figure 7 3 Find WWPN in ESXi 2 Create at least one FB volume 3 Create a volume group using hosttype VMware and assign the volume s to the group Define Volume Group Properties Define the volume group properties A volume group is a set of logical volumes that can be accessed by a host Volume Group Nickname Host Type new_esxi_va VMWare ESX VMWare A Figure 7 4 DS Storage Manager GUI Select hosttype for volume group 4 Create host connections using hosttype VMware Define Host Ports Define one or more host ports that you will use to map hosts to a volume group in the next step After you add the table is updated All of the host ports in the table will be mapped to the same volume group when you crea Host Nickname Port Type esxi_host Fibre Channel Point to Point Switched FcSf e Host Type VMWare ESX VMWare Figure 7 5 DS Storage Manager GUI Select hosttype for host connection IBM System Storage DS8000 Host Attachment and Interoperability 5 Map the volume group to the host 6 In the vSphere client go to Configuration gt Storage adapters and click Rescan All in the top right corner see Figure 7 6 Storage Adapters Add Remove Refresh Device Type 82801G ICH7 Family IDE Controller _ vmhba32 Block SCSI ISP81xx based 10 GbE FCoE to PCI Express CNA vmhba6 Fibre Channel 20 00 00 c0 dd 18 3b 79 21 00 00 c0 dd 18 3b 79 vmhba7 Fibre C
55. HYPERPAV ALIAS 266 IBM System Storage DS8000 Host Attachment and Interoperability The D M DEV command in Example 12 10 shows HyperPAV aliases HA Example 12 10 The system configuration information shows the HyperPAV aliases SY1 d m dev SY1 1IEE174I 23 42 09 DISPLAY M 844 DEVICE STATUS NUMBER OF ONLINE CHANNEL PATHS 0 1 2 3 4 567 8 9 ABC OD E F 000 DN 4 DN DN DN DN DN DN DN DNDN1 1 1 1 018 DN DN DN DN 4 DN DN DN DN DN DN DN DN DN DN DN 02E 4 DN4 DN4 8 4 4 4 4 4 4 4 DN4 ON 02F DN4 4 4 4 4 4 DN4 4 4 4 4 ODN DN 4 030 8 0334 tis g glen chee Sh Ser Geo Whe Se Gad tes ee 034 4 4 4 4 ODN DN DN DN DN DN DN DN DN DN DN DN 03E 1 DN DN DN DN DN DN DN DN DN DN DN DN DN DN DN 0414 4 4 4 4 4 4 4 AL AL AL AL AL AL AL AL 048 4 4 ODN DN DN DN DN DN DN DN DN DN DN DN DN 4 0514 4 4 4 4 4 4 4 UL UL UL UL UL UL UL UL 0614 4 4 4 4 4 4 4 AL AL AL AL AL AL AL AL 071 4 4 4 4 DN DN DN DN HA HA DNDN 073 DN DN DN DN DN DN DN HA HA 098 4 4 4 4 N8 4 4 4 4 4 DN4 4 4 4 OEO DN DN 1 DN DN DN DN DN DN DN DN DN DN DN DN DN OF1 1 DN DN DN DN DN DN DN DN DN DN DN DN DN DN DN FFF HA HA HA HA ce cerdelerekenueekeiant SYMBOL EXPLANATIONS kkkkkkkkkkkkkkkkkkkkkkkk ONLINE PHYSICALLY ONLINE AND OPERATIONAL INDICATORS ARE NOT EQUAL ONLINE DEVICE OFFLINE DOES NOT EXIST BX DEVICE IS BOXED SN SUBCHANNEL NOT AVAILABLE DN DEVICE NOT AVAILABLE PE SUBCHANNEL IN PERMANENT ERROR AL DEVICE IS AN ALIAS UL DEVICE IS AN UNBOUND ALIAS HA
56. I O in failover mode only By changing this value for selected paths to 2 or even a higher value the path priority is lowered and only the highest priority path is going to be used for I O in failover mode as long as it is available Example 5 8 shows the set the priority of the second path through the adapter vscsi1 connected to VIOS 2 of hdisk2 to 2 so that the first path through the adapter vscsi0 connected to VIOS 1 with a priority of 1 remains the preferred path for I O By applying this pattern for each of the available virtual SCSI disks devices in the partition in a load balanced rotating method I O can manually be balanced across both VIOS partitions Example 5 8 Using the AIX chpath command to balance I O across paths in a dual VIOS setup Ispath 1 hdisk0 Enabled hdisk0 vscsi0 Enabled hdisk0O vscsil Ispath AHE 1 hdiskO p vscsi0 attribute value description user settable priority 1 Priority True Ispath AHE 1 hdiskO p vscsil attribute value description user settable priority 1 Priority True chpath 1 hdiskO p vscsil a priority 2 path Changed Ispath AHE 1 hdiskO p vscsil attribute value description user settable priority 2 Priority True For AIX MPIO multipath support of virtual SCSI devices in the AIX client partition the DS8000 LUN must be presented as a physical device hdiskx from the VIOS to the client partition It is not possible to provide a single DS8000 LUN and then further subdivide it into
57. IT Specialist and Project Leader for IBM System Storage disk products at the International Technical Support Organization San Jose Center He has worked at IBM in various IT areas Bertrand has written many IBM Redbooks publications and has also developed and taught technical workshops Before joining the ITSO he worked for IBM Global Services as an Application Architect in the retail banking telecommunication and health care industries He holds a Masters degree in Electrical Engineering Juan Brandenburg is a Product Field Engineer for the DS8000 in the USA He is a graduate from the University of Arizona holding a Bachelors of Engineer Management in Computer Engineering His areas of experience for hardware include the DS8000 series and IBM System x series server Juan has been working for IBM for seven years in the Global Mirror ESSNI and DS8000 departments He has many years of experience scripting for Linux IBM AIX and Windows environments Juan has continuously participated in IBM technical competitions for which he has earned him some awards such as the Distinguished Engineer Award for the 2006 Tech Connect competition placing him in the IBM Epic Express top 14 coops for July 2006 Jana Jamsek is an IT Specialist for IBM Slovenia She works in Storage Advanced Technical Support for Europe as a specialist for IBM Storage Systems and the IBM i formerly known as the IBM i5 OS operating system Jana has eight years of experience
58. Interoperability device vendor TBM product 2107900 path_grouping policy group_by_serial User friendly setting Settings for the user_friendly_names parameter are provided in Example 6 21 on page 99 You can leave it as it is or remove it from the file The values for polling interval failback no_path_retry and path_checker control the behavior of DM Multipath in case of path failures For most instances these values should not be changed If you need to modify these parameters see 6 1 2 Attaching Linux server to DS8000 resources on page 82 for more information The rr_min_io setting specifies the number of I O requests that are sent to one path before switching to the next one In certain cases you can improve performance by lowering this value from 100 to 32 Important In SLES 11 SP1 rr_min_io does not work in the defaults section Move it to the devices section for example devices device vendor TBM product 2107900 path grouping policy group_by_serial rr_min_io 32 System z multipath settings For Linux on System z multipathing set the dev_loss_tmo parameter to 90 seconds and the fast_io_fail_tmo parameter to 5 seconds Modify the etc multipath conf file to specify the values as shown in Example 6 34 Example 6 34 System z multipath settings defaults dev_loss_tmo 90 fast_io_fail_tmo 5 Make the changes effective by using the reconfigure command in the interactive multipathd k pr
59. LUNs Therefore boot or IPL from SAN is a prerequisite for this function Chapter 11 IBM i considerations 243 Why consider cloning By using the cloning capability you can create a complete copy of your entire system in minutes You can then use this copy in any way you want for example you can use it to minimize your backup windows or protect yourself from a failure during an upgrade or even use it as a fast way to provide yourself with a backup or test system You can do all of these tasks using cloning with minimal impact to your production operations When to use cloning You might want to use cloning in the following circumstances gt You need enough free capacity on your external storage unit to accommodate the clone If Metro Mirror or Global Mirror are used you need enough bandwidth on the links between the DS8000 primary and secondary site gt Consider that the copy services need enough resources on the primary and secondary DS8000 to ensure that there will be only a minimal performance impact on the production IBM i system gt Do not attach a clone to your network until you have resolved any potential conflicts that the clone has with the parent system 11 9 Migrating For many System i clients migrating to the DS8000 is best achieved by using traditional save and restore techniques However several alternatives are available to consider For more information see 11 9 1 Metro Mirror and Global Copy on p
60. Licensed Function that must be activated before you can start to use it For details about how to obtain the necessary information for the DSFA website and how to apply the license keys see the Redbooks publication IBM System Storage DS8000 Architecture and Implementation SG24 8886 In the Workload Manager WLM Service Definition the service option I O priority management must be set to Yes in order to be able to use it See also Figure 12 6 on page 277 to verify where it can be enabled on the DS8000 Chapter 12 IBM System z considerations 275 Support is provided for CKD I O priority with software input gt The user assigns a performance policy to each CKD volume that applies in the absence of additional software support gt The OS can optionally specify parameters that determine priority of each I O operation gt In general the priorities will be managed by z OS through eWLM gt Itis supported on z OS V1 11 V1 12 V1 13 and above An SMNP trap is automatically generated when a rank enters saturation gt Without z OS software support On rank saturation volume I O is managed according to volumes performances groups performance policy defined into HMC gt With z OS software support The user assigns applications priorities through eWLM WorkLoad Manager z OS assigns an importance value to each I O based on eWLM output z OS assigns an achievement value to each I O based on prior history I O respons
61. Listing Volumes LSS Volume Size Assigned to 10 OO1AAGXA 1 00000072E10 BytesUnassigned 10 003AAGXA 5 3687091E9 Bytes5000000000000000 11 103AAGXA 2 14748365E10 Bytes5000000000000000 Chapter 3 Windows considerations 35 3 10 Hyper V considerations This section provides information for the specifics of attaching IBM System Storage DS8000 systems to host systems running Microsoft Hyper V Server 2008 R2 This section includes the following topics gt Yv y Hyper V introduction Storage concepts for virtual machines Cluster Shared Volumes CSV Best practices 3 10 1 Hyper V introduction 36 Hyper V is Microsoft platform for x86 based server virtualization Hyper V exists in two versions gt As Microsoft Hyper V Server 2008 R2 This version provides a core installation of Microsoft Windows Server 2008 with Hyper V functionality Other Windows Server roles are not included and it comes with limited Windows Services for management This version is free and can be downloaded at the following website http www microsoft com en us download details aspx id 20196 As a server role in Microsoft Windows Server 2008 R2 Hyper V can be added as a server role in Windows Server 2008 R2 It provides the same virtualization functionality as the plain Hyper V Server but also provides a full Windows Server 2008 including other server roles such as web server Domain controller and so on and a build in graphical management
62. NORMAL 0 0 2 fscsil path2 CLOSE NORMAL 0 0 3 fscsil path3 CLOSE NORMAL 0 0 DEV 13 DEVICE NAME hdisk13 TYPE 2107900 ALGORITHM Load Balance SERIAL 75TV1816201 Path Adapter Path Name State Mode Select Errors 0 fscsi0 path0 CLOSE NORMAL 0 0 1 fscsi0 path1 CLOSE NORMAL 0 0 2 fscsil path2 CLOSE NORMAL 0 0 3 fscsil path3 CLOSE NORMAL 0 0 This method significantly reduces number of devices in the system see Example 5 19 and helps with managing devices For example when removing a physical volume from a system using rmdev dl hdiskx is sufficient Example 5 19 Isdev command Isdev Cc disk tests hdisk12 Availabl hdisk13 Availabl hdisk14 Availabl hdisk15 Availabl hdisk16 Availabl hdisk17 Availabl hdisk18 Availabl hdisk19 Availabl hdisk20 Availabl hdisk21 Availabl 00 08 02 IBM MPIO FC 2107 00 08 02 IBM MPIO FC 2107 00 08 02 IBM MPIO FC 2107 00 08 02 IBM MPIO FC 2107 00 08 02 IBM MPIO FC 2107 00 08 02 IBM MPIO FC 2107 00 08 02 IBM MPIO FC 2107 00 08 02 IBM MPIO FC 2107 00 08 02 IBM MPIO FC 2107 00 08 02 IBM MPIO FC 2107 ooo 0 0 OD OD MO MO MO For path management you can use AIX built in commands such as chpath rmpath or you might find the SDDPCM pcmpath command more flexible Starting from SDDPCM version 2 4 0 3 a new device attribute retry_timeout is added for ESS DS6K DS8K SVC devices This attribute allows the user to set the time out value for I O retry on the last path The default value of
63. NPIV With regard to continuous availability for virtual I O typically two VIOS partitions are deployed on a managed system to provide highly available virtual SCSI virtual Fibre Channel and shared Ethernet services to client partitions The main reason to implement two VIOS servers dual VIOS setup is to provide continuous access when a VIOS must be taken offline for planned outages such as software updates 4 2 Using VSCSI with IBM VIOS and DS8000 VIOS enables virtualization of physical storage resources using generic SCSI block level emulation Virtualized storage devices are accessed by the client partitions through virtual SCSI devices All virtual SCSI devices are accessed as standard SCSI compliant LUNs by the client partition Virtual SCSI enables client logical partitions LPARs to share disk storage and tape or optical devices that are assigned to the VIOS logical partition The functionality for virtual SCSI is provided by the IBM POWER Hypervisor Virtual SCSI helps enable secure communications between partitions and a VIOS that provides storage backing devices Disk tape or optical devices attached to physical adapters in the VIOS logical partition can be shared by one or more client logical partitions The VIOS acts as a standard storage subsystem that provides standard SCSI compliant LUNs The VIOS is capable of exporting a pool of heterogeneous physical storage as a homogeneous pool of block storage in the form of SCSI dis
64. PartO TYPE 2107900 POLICY OPTIMIZED SERIAL 75207814703 Path Adapter Hard Disk State Mode Select Errors 0 Scsi Port1 BusO Disk2 Part0 OPEN NORMAL 203 4 1 Scsi Port2 BusO Disk2 Part0 OPEN NORMAL 173 1 DEV 1 DEVICE NAME Disk3 PartO TYPE 2107900 POLICY OPTIMIZED SERIAL 75ABTV14703 Path Adapter Hard Disk State Mode Select Errors 0 Scsi Port1 Bus0 Disk3 Part0 OPEN NORMAL 180 0 1 Scsi Port2 Bus0 Disk3 Part0 OPEN NORMAL 158 0 DEV 2 DEVICE NAME Disk4 PartO TYPE 2107900 POLICY OPTIMIZED SERIAL 75034614703 Path Adapter Hard Disk State Mode Select Errors 0 Scsi Port1 BusO Disk4 Part0 OPEN NORMAL 221 0 1 Scsi Port2 Bus0 Disk4 Part0 OPEN NORMAL 159 0 All paths of the DS8000 will be used because the DS8000 does not implement the concept of a preferred path Additionally the datapath query wwpn command Example 3 4 shows the WWPN of your Fibre Channel adapter Example 3 4 The datapath query wwpn command with SDDDSM C Program Files IBM SDDDSM gt datapath query wwpn Adapter Name PortWWN Scsi Port 210000E08B1EAE9B Scsi Port2 210000E08B0B8836 The datapath query portmap command provides a map of the DS8000 I O ports and shows on which I O ports your HBAs are connected see Example 3 5 Example 3 5 The datapath query portmap command C Program Files IBM SDDDSM gt datapath query portmap ESS 75207 7 5ABT 75034 v1 ZO KX no BAY 1 B1 BAY 2 B2 BAY 3 B3 BAY 4 B4 ID DISK H1 H2 H3 H4 H1 H2 H3 H4 H1 H2 H3 H4 H1 H
65. Systems the VIOS enables sharing of physical resources between logical partitions LPARs including virtual SCSI virtual Fibre Channel using node port ID virtualization NPIV and virtual Ethernet providing a more efficient utilization of physical resources and facilitating server consolidation VIOS is part of the IBM PowerVM editions hardware feature on IBM Power Systems The VIOS technology facilitates the consolidation of network and disk I O resources and minimizes the number of required physical adapters in the IBM Power Systems server It is a special purpose partition that provides virtual I O resources to its client partitions The VIOS actually owns the physical resources that are shared with clients A physical adapter assigned to the VIOS partition can be used by one or more other partitions VIOS can provide virtualized storage devices storage adapters and network adapters to client partitions running an AIX IBM i or Linux operating environment The VIOS includes the following core I O virtualization capabilities gt Virtual SCSI gt Virtual Fibre Channel using NPIV gt Virtual Ethernet bridge using a shared Ethernet adapter SEA The core I O virtualization capabilities are represented in Figure 4 1 AIX 5 3 AIX 6 1 AIX 5 3 AIX 6 1 AIX 5 3 AIX 6 1 AIX 7 1 Power i AIX 7 1 Power i AIX 7 1 Poweri or Linux Siwy AN Virtual SCSI Ethernet a it Bridge irae i ee POWER Hypervisor
66. The I O supervisor IOS recovery was designed to support a small number of devices per control unit and a unit check was presented on all devices at failover It does not scale well with a DS8000 which has the capability to scale up to 65 280 devices Under these circumstances you can have CPU or spinlock contention or exhausted storage under the 16 M line at device failover or both Starting with z OS V1 4 and higher for DS8000 software support the IOS recovery is improved by consolidating unit checks at an LSS level instead of each disconnected device This consolidation shortens the recovery time as a result of I O errors This enhancement is particularly important because the DS8000 can have up to 65 280 devices in a storage facility Scalability benefits With enhanced scalability support the following benefits are possible gt Common storage area CSA usage above and under the 16 M line is reduced gt The IOS large block pool for error recovery processing and attention and the state change interrupt processing are located above the 16 M line thus reducing the storage demand under the 16 M line gt Unit control blocks UCB are pinned during event notification facility signalling during channel path recovery Chapter 12 IBM System z considerations 249 gt Scalability enhancements provide additional performance improvements as follows Bypassing dynamic pathing validation in channel recovery for reduced recovery
67. admin diskutil list dev disk0 TYPE NAME SIZE IDENTIFIER 0 GUID_partition_scheme 82 0 GB disk0 1 EFI 209 7 MB diskOsl 2 Apple_HFS HDD 34 0 GB disk0s2 3 Apple_HFS HDD 2 47 5 GB disk0s3 dev diskl TYPE NAME SIZE IDENTIFIER 0 GUID_partition_scheme 128 8 GB diskl l EFI 209 7 MB diskisl 2 Apple_HFS DS8k02 128 5 GB diskls2 dev disk2 TYPE NAME SIZE IDENTIFIER 0 GUID_partition_scheme 107 4 GB disk2 I EFI 209 7 MB disk2s1 2 Apple_HFS DS8k01 107 0 GB disk2s2 The system manpages man diskutil should be referenced for detailed information and examples of using the diskuti command 8 6 Troubleshooting The ATTO Technology Inc webpage contains very useful information for troubleshooting Website http attotech com On the ATTO web site click support gt Troubleshooting Tips amp FAQs Celerity Fibre Channel Host Adapters gt Troubleshooting Celerity Fibre Channel Host Adapters Mac 8 6 1 Useful Utilities on Apple Mac OSX Disk Utility OS X disk repair and formatting tool Location Applications Utilities DiskUtility Chapter 8 Apple considerations 157 Console Allows the user to see application logs in real time The ATTO Celerity FC Driver does not use the system log to report events Location Applications Utilities Console Apple System Profiler This System Utility shows useful system information All displayed information can be saved in a file Location Apple Menu gt A
68. balancing The path chosen based on estimated path load Default policy The policy can be set through the datapath set device policy command 9 4 2 Using MPxlO On Solaris 8 and Solaris 9 systems MPxIO is available as a bundling of OS patches and packages You must install these patches and packages to use MPxIO On Solaris 10 systems MPxIO is installed by default In all cases MPxIO needs to be enabled and configured before it can be used with the DS8000 Important Although widely assumed it is not a must to use Sun badged HBAs to run MPXxIO The native HBAs from vendors such as Emulex and QLogic can be used as well For more details see the following websites gt For Emulex http www oracle com technetwork server storage solaris overview emulex cor poration 136533 htm gt For Qlogic http www oracle com technetwork server storage solaris overview qlogic cor p 139073 htm Before you enable MPxIO configure the host bus adapters Issue the cfgadm 1a command for the current state of your adapters Example 9 6 shows two adapters c3 and c4 of type fc Example 9 6 cfgadm la command output cfgadm la Ap_Id Type Receptacle Occupant Condition c3 fc connected unconfigured unknown c4 fc connected unconfigured unknown The command output shows that both adapters are unconfigured To configure the adapters issue cfgadm c configure cx where x is the adapter number in this case 3 and 4 Now both adapters
69. be booted off that DS8000 LUN afterwards Example 9 22 Solaris that was booted off a DS8000 LUN bash 3 00 df dev dsk c5t600507630AFFC29F0000000000003101d0s0 3667996 blocks 564069 files devices devices 0 blocks 0 files system contract ctfs 0 blocks 2147483621 files proc proc 0 blocks 29933 files etc mnttab mnttab 0 blocks 0 files etc svc volatile swap 6904240 blocks 558165 files system object objfs 0 blocks 2147483497 files etc dfs sharetab sharefs 0 blocks 2147483646 files dev fd fd 0 blocks 0 files tmp swap 6904240 blocks 558165 files var run swap 6904240 blocks 558165 files export home dev dsk c5t600507630AFFC29F0000000000003101d0s7 3263196 blocks 407676 files bash 3 00 format Searching for disks done AVAILABLE DISK SELECTIONS 0 cOtOd0 lt FUJITSU MAW3073NC 3701 cyl 24345 alt 2 hd 8 sec 737 gt pci 83 4000 FUSV ulsa 2 1 sd 0 0 1 clt0d0 lt FUJITSU MAW3073NC 3701 cyl 24345 alt 2 hd 8 sec 737 gt pci 83 4000 FISV ulsa 2 sde0 0 2 c5t600507630AFFC29F0000000000003100d0 lt IBM 2107900 268 cyl 4367 alt 2 hd 30 sec 64 gt scsi_vhci ssd g600507630af fc29f0000000000003100 3 c5t600507630AFFC29F0000000000003101d0 lt IBM 2107900 268 cyl 8736 alt 2 hd 30 sec 64 gt scsi_vhci ssd g600507630af fc29f0000000000003101 4 c5t600507630AFFC29F0000000000003102d0 lt IBM 2107900 268 cyl 17918 alt 2 hd 64 sec 256 gt scsi_vhci ssd g600507630af fc29f0000000000003102 174 IBM Sys
70. be provided through the Application Migration Assistance Tracker For more details about Assistance Tracker see APAR 1113752 at this website http www ibm com support docview wss uid isg1 1113752 gt Actions advised Review your programs and take a look at the calls for the macros OBTAIN REALLOC CVAFDIR CVAFSEQ CVAFDSM and CVAFFILT These macros were changed and you need to update your program to reflect those changes IBM System Storage DS8000 Host Attachment and Interoperability Look for programs that calculate volume or data set size by any means including reading a VTOC or VTOC index directly with a basic sequential access method or execute channel program EXCP data control block DCB This task is important because now you will have new values returning for the volume size Review your programs and look for EXCP and Start I O STARTIO macros for DASD channel programs and other programs that examine DASD channel programs or track addresses Now that you have the new addressing mode your programs must be updated Look for programs that examine any of the many operator messages that contain a DASD track block address data set or volume size The messages now show new values gt Migrating data Define new EAVs by creating them on the DS8000 or expanding existing volumes using dynamic volume expansion Add new EAV volumes to storage groups and storage pools and update automatic class sele
71. channel impact In older storage disk systems a device had an implicit allegiance that is a relationship created in the control unit between the device and a channel path group when an I O operation is accepted by the device The allegiance causes the control unit to guarantee access no busy status presented to the device for the remainder of the channel program over the set of paths associated with the allegiance With Multiple Allegiance the I O requests are accepted by the DS8000 and all requests are processed in parallel unless there is a conflict when writing to the same data portion of the CKD logical volume In systems without Multiple Allegiance all except the first I O request to a shared volume are rejected and the I Os are queued in the System z channel subsystem showing up in Device Busy Delay and PEND time in the RMF DASD Activity reports However a device busy condition can still happen It will occur when an active I O is writing a certain data portion on the volume and another I O request comes in and tries to either read or write to that same data To ensure data integrity those subsequent I Os will get a busy condition until that previous I O is finished with the write operation Multiple Allegiance provides significant benefits for environments running a sysplex or System z systems sharing access to data volumes Multiple Allegiance and PAV Parallel Access Volumes can operate together to handle multiple requests from
72. copy 01 offset 000000 enabled config priv 000256 048207 047952 copy 01 offset 000192 enabled log priv 048208 055503 007296 copy 01 offset 000000 enabled lockrgn priv 055504 055647 000144 part 00 offset 000000 Multipathing information numpaths 2 c2t50050763030CC1A5d0s2 state enabled c3t50050763031CC1A5d0s2 state enabled In Example 9 12 the capacity of the disk as shown in VxVM can be found on the output line labeled public after issuing a vxdisk list lt dmpnodename gt command Multiply the value for len by 512 bytes which is equal to 12 GB 25095808 x 512 168 IBM System Storage DS8000 Host Attachment and Interoperability After the file system on the logical volume of the VxVM diskgroup is created the size of the file system 11 GB is mounted on 20781 and displayed using the df command Example 9 13 Example 9 13 The df command before volume expansion bash 3 00 df k Filesystem kBytes used avail capacity Mounted on dev dsk c1t0d0s0 14112721 4456706 9514888 32 devices 0 0 0 0 devices ctfs 0 0 0 0 system contract proc 0 0 0 0 proc mnttab 0 0 0 0 etc mnttab swap 7225480 1160 7224320 1 etc svc volatile objfs 0 0 0 0 system object fd 0 0 0 0 dev fd swap 7224320 0 7224320 0 tmp swap 7224368 48 7224320 1 var run swap 7224320 0 7224320 0 dev vx dmp swap 7224320 0 7224320 0 dev vx rdmp dev dsk c1t0d0s7 20160418 2061226 17897588 11 export home dev vx dsk 03461_dg 03461_vol 104
73. dev deviceFileSystem Usb MassStorage dsk disk hp 1008 294 A60020000001 dev deviceFileSystem Usb MassStorage rdsk disk hp 1008 294 A60020000001 disk 6 0 3 1 0 1 12 255 0 0 0 sdisk CLAIMED DEVICE IBM 2107900 disk 7 0 3 1 0 1 12 255 0 0 1 sdisk CLAIMED DEVICE IBM 2107900 disk 8 0 3 1 0 1 13 255 0 0 0 sdisk CLAIMED DEVICE IBM 2107900 disk 9 0 3 1 0 1 13 255 0 0 1 sdisk CLAIMED DEVICE IBM 2107900 disk 4 0 3 1 0 1 14 255 0 0 0 sdisk CLAIMED DEVICE IBM 2107900 disk 5 0 3 1 0 1 14 255 0 0 1 sdisk CLAIMED DEVICE IBM 2107900 disk 2 0 3 1 0 1 15 255 0 0 0 sdisk CLAIMED DEVICE IBM 2107900 disk 3 0 3 1 0 1 15 255 0 0 1 sdisk CLAIMED DEVICE IBM 2107900 disk 10 0 4 1 0 0 0 0 0 sdisk CLAIMED DEVICE HP DHO72ABAA6 Chapter 10 HP UX considerations 185 dev dsk c30t0d0 dev dsk c30t0d0s2 dev rdsk c30t0d0 dev dsk c30t0d0s1 dev dsk c30t0d0s3 dev rdsk c30t0d0s1 disk 11 0 4 1 0 0 0 1 0 sdisk CLAIMED DEVICE dev dsk c30t1d0 dev rdsk c30t1d0 HP dev rdsk c30t0d0s2 dev rdsk c30t0d0s3 DHO72ABAA6 There are two options to create the missing special device file The first option is to restart the host which is disruptive The alternative is to run the command insf eC disk which will reinstall the special device files for all devices of the class disk After creating the special device files the ioscan command displays discovered DS8000 devices as shown in Example 10 5 Example 10 5 Discovered DS8000 devices with a special device file
74. dev disk by id directory are not changed These nodes still exist and have the volumes unique IDs in their names Accessing DM Multipath devices in Red Hat Enterprise Linux 5 Red Hat Enterprise Linux sets the user_friendly_names option in its default etc multipath conf file The devices it creates in the dev mapper directory is shown in Example 6 31 Example 6 31 Multipath devices in Red Hat Enterprise Linux 5 in dev mapper root x36501ab9 1s 1 dev mapper cut c 45 mpathl mpath2 mpath3 mpath4 A second set of device nodes contain the unique IDs of the volumes in their name whether user friendly names are specified or not These nodes are in the dev mpath directory Using multipath devices Use the device nodes that are created for multipath devices just like any other block device gt Create a file system and mount it gt Use LVM on the device nodes gt Build software RAID devices Chapter 6 Linux considerations 103 You can also partition a DM Multipath device using the fdisk command or any other partitioning tool To make new partitions on DM Multipath devices available use the partprobe command It triggers udev to set up new block device nodes for the partitions as shown in Example 6 32 Example 6 32 Using the partprobe command to register newly created partitions x36501ab9 fdisk dev mapper 3600507630af fc29f0000000000004691 lt all steps to create a partition and write the new partition tabl
75. developing using marketing or distributing application programs conforming to the application programming interface for the operating platform for which the sample programs are written These examples have not been thoroughly tested under all conditions IBM therefore cannot guarantee or imply reliability serviceability or function of these programs Copyright IBM Corp 2011 2013 All rights reserved ix Trademarks IBM the IBM logo and ibm com are trademarks or registered trademarks of International Business Machines Corporation in the United States other countries or both These and other IBM trademarked terms are marked on their first occurrence in this information with the appropriate symbol or indicating US registered or common law trademarks owned by IBM at the time this information was published Such trademarks may also be registered or common law trademarks in other countries A current list of IBM trademarks is available on the Web at http www ibm com legal copytrade shtml The following terms are trademarks of the International Business Machines Corporation in the United States other countries or both AIX Lotus Storwize BladeCenter MVS System i CICS Notes System p DB2 Power Architecture System Storage DS developerWorks POWER Hypervisor System Storage DS8000 Power Systems System x Dynamic Infrastructure POWER6 System z10 Easy Tier POWER6 System z ECKD
76. disks to the DS8000 whether the existing disks are internal or external When you assign the new DS8000 logical volumes to the System i initially they are non configured For more information see 11 6 8 Adding volumes to the System i configuration on page 224 IBM System Storage DS8000 Host Attachment and Interoperability If you add the new units and choose to spread data IBM i automatically migrates data from the existing disks onto the new logical units You can then use the STRASPBAL TYPE ENDALC IBM i command to mark the units to remove from the configuration as shown in Figure 11 40 This command can help reduce the downtime associated with removing a disk unit and keeps new allocations away from the marked units Start ASP Balance STRASPBAL Type choices press Enter Balance type gt ENDALC CAPACITY USAGE HSM Storage unit 1 4094 for more values Bottom F3 Exit F4 Prompt F5 Refresh F12 Cancel F13 How to use this display F24 More keys Figure 11 40 Ending allocation for existing disk units When you later run the STRASPBAL TYPE MOVDTA IBM i command all data is moved from the marked units to other units in the same ASP as shown in Figure 11 41 You must have sufficient new capacity to allow the data to be migrated Start ASP Balance STRASPBAL Type choices press Enter Balance type gt MOVDTA CAPACITY USAGE HSM Time limit 1 999
77. empty and that no data will be overwritten Click Next E Disk LUN Select Disk LUN File System Version Device Capa A Current Disk Layout IBM Fibre Channel Disk naa Non SSD 20 00 GB 20 00 GB Properties cation Formatting vmfs devices disks naa 600507630affc29f000000000000a000 Ready to Complete Partition F at Review the current disk layout GPT The hard disk is blank There is only one layout configuration available Use the Next button to proceed with the othe Pages A partition will be created and used Figure 7 12 Current disk layout 128 IBM System Storage DS8000 Host Attachment and Interoperability e Enter a name for the new datastore and click Next Figure 7 13 E Disk LUN m Enter a datastore name Select Disk LUN File System Version DS8000_Datastore 1 Current Disk Layout Properties Formatting Ready to Complete Figure 7 13 Datastore name f Select the size of the new datastore from the page shown in Figure 7 14 Click Next Tip Always use the full size of the LUN Having multiple datastores or partitions on one LUN is not a good practice E Disk LUN x Capacity a Select Disk LUN File System Version Maximum available space Current Disk Layout C Custom space setting am anil 20 00 GB of 20 00 GB available space Ready to Complete Figure 7 14 Datastore size g Check the summary in the Ready to Complete window shown in Figure 7 15 and click F
78. entl U9117 MMA O6C6DE1 V2 C20 Virtual I O Slot vfchost0 U9117 MMA O6C6DE1 V2 C21 Virtual I O Slot vfchostl U9117 MMA O6C6DE1 V2 C22 Virtual I O Slot vfchost2 U9117 MMA 06C6DE1 V2 C23 Virtual I O Slot vfchost3 U9117 MMA O6C6DE1 V2 C24 Virtual I O Slot vfchost10 U9117 MMA 06C6DE1 V2 C25 Virtual I O Slot vfchost11 U9117 MMA 06C6DE1 V2 C30 Virtual I O Slot vfchost4 U9117 MMA 06C6DE1 V2 C31 Virtual I 0 Slot vfchost5 U9117 MMA 06C6DE1 V2 C32 Virtual I 0 Slot vfchost6 U9117 MMA 06C6DE1 V2 C33 Virtual I O Slot vfchost7 U9117 MMA 06C6DE1 V2 C34 Virtual I O Slot vfchost12 U9117 MMA 06C6DE1 V2 C35 Virtual I 0 Slot vfchost13 U9117 MMA O6C6DE1 V2 C70 Virtual I O Slot vfchost8 U9117 MMA 06C6DE1 V2 C71 Virtual I 0 Slot vfchost9 Figure 11 33 Virtual devices in VIOS Before installation the virtual adapters still report a status of NOT _LOGGED_IN Figure 11 34 Ismap vadapter vfchost12 npiv Name Physloc ClntID ClntName CIntOS vfchost12 U9117 MMA 06C6DE1 V2 C34 19 Status NOT_LOGGED_IN FC name fcs2 FC loc code U789D 001 DQD904G P1 C6 T1 Ports logged in 0 Flags 4 lt NOT_LOGGED gt VFC client name VFC client DRC Ismap vadapter vfchost13 npiv Name Physloc ClntID ClntName CIntOS vfchost13 U9117 MMA 06C6DE1 V2 C35 19 Status NOT _LOGGED_IN FC name fcs3 FC loc code U789D 001 DQD904G P1 C6 T2 Ports logged in 0 Flags 4 lt NOT_LOGGED gt VFC client name VFC client DRC Figure 11 34 Reporting of virtual Fibre
79. extent device ayou Complete IBM Fibre Channel Disk naa 600 Non SSD 40 00 GB 1 vmfs devices disks naa 600507630affc29f000000000000a000 GPT Primary Partitions Free space 1 00 MB VMFS IBM Fibre Channel Disk naa 40 00 GB The following VMFS will be increased as shown Properties Extents Datastore name DS8000_Datastore1 IBM Fibre Channel Disk naa 600507630 Fi tti Total capacity 40 00 GB File system vmfs 5 Block size 1MB Maximum file size 2 00 TB Figure 7 24 Summary page h Close the Datastore Properties window 3 Verify from the Storage view that the datastore has successfully been expanded 134 IBM System Storage DS8000 Host Attachment and Interoperability 7 2 2 Virtual disks The most common concept used with ESXi are virtual disks Virtual disk concepts From an ESXi view virtual disks are large files that represent the virtual hard drives Due to the maximum file size of 2 TB on VMFS a virtual disk can have a maximum size of 2 TB Inside a VM a virtual disk looks like a SCSI hard drive All SCSI I O commands issued by the operating system inside the VM are converted into file I O operations by the hypervisor This processing is transparent to the operating system Assigning a virtual disk to a VM To assign a virtual disk to a VM perform the following steps 1 Select the VM from the list right click and select Edit Settings as in Figure 7 25 SUES AA Virtual Machines lf l
80. for MBps per port is based on the measurements of maximum sequential and transactional throughput of a port Based on DS8000 reports from typical workloads a rough estimation is made that about 20 of the IBM i I Os are sequential Accordingly the calculation is made for the weighted average between transaction MBps and sequential MBps and applied 70 for port utilization Chapter 11 IBM i considerations 213 When planning for multipath consider that two or more ports are used for I O Therefore you can implement two times or more times the preferred capacity per port Table 11 6 Sizing for Fibre Channel adapters in IBM i IOP Adapter port IOPS at 70 Disk MBps utilization capacityper per port port GB 2844 5760 3 200 853 45 MBps 4 Gb IOP based Fibre Channel adapter Port is 5749 5774 10 500 2 800 183 MBps 4 Gb IOP less Fibre Channel adapters Port in 5735 12 250 3 267 202 MBps 8 Gb IOP less Fibre Channel adapter Example You plan to implement DS8000 connected to IBM i with 8 Gb adapters feature number 5735 You plan for the LUNs of capacity 141 GB By the performance guideline in Table 11 6 you consider 3267 GB per port so you might want to connect 3267 141 23 LUNs per port or 46 LUNs per 2 ports in 2 different adapters for multipath Fibre Channel adapters in VIOS When connecting the DS8000 series to IBM i with VIOS NPIV or VIOS VSCSI follow the guidelines in Table 11 7 to
81. for less alias device addresses and for more real or base device addresses within a LCU RMF considerations RMF reports all I O activity against the base PAV address not by the base and associated aliases The performance information for the base includes all base and alias activity As illustrated in Example 12 6 on the Data Set Delays window in RMF Monitor Ill for Volume MLDC65 with device Number DC65 you can see the identification for an EAV volume 3390A in the Device field The presence of PAV H means that HyperPAV is used Example 12 6 Data Set Delays Volume window RMF V1R10 Data Set Delays Volume Command gt Scroll gt CSR Samples 120 System SC70 Date 05 10 09 Time 18 04 00 Range 120 Sec Volume MLDC65 Device Data Number DC65 Active 0 Pending 0 Average Users Device 3390A Connect 0 Delay DB 0 Delayed Shared Yes Disconnect 0 Delay CM 0 0 0 PAV 1 0H Chapter 12 IBM System z considerations 263 Data Set Name Jobname ASID DUSG DDLY No I 0 activity detected Missing interrupt handler values considerations The DS8000 provides a preferred interval of 30 seconds as part of the RCD The z OS uses this information to set its missing interrupt handler MIH value The MIH times for PAV alias addresses must not be set An alias device inherits the MIH of the base address to which it is bound and it is not
82. hard disk to the guest operating system Select the type of disk to use Disk Create a new virtual disk Use an existing virtual disk Reuse a previously configured virtual disk Raw Device Mappings Give your virtual machine direct access to SAN This option allows you to use existing SAN commands to manage the storage and continue to access it using a datastore Figure 7 32 Raw Device Mapping 5 Select the LUN to use as RDM from the list as shown in Figure 7 33 Name Identifier Path ID LUN or Capacity contains v Clear Path ID LUN Capacity Hare vmhba2 C0 T0 12 2 25 00GB Unk Name IBM Fibre Channel Disk naa 60050 Figure 7 33 Select LUN for the RDM 6 Select where to store the mapping file for the RDM as shown in Figure 7 34 Select the datastore on which to store the LUN mapping You will use the disk map on this datastore to access the virtual disk Store with Virtual Machine C Specify datastore Datastore Hosts Datastore Cluster Local_Disk 1 N A DS8000_Datastore1 1 N A Figure 7 34 Datastore for the RDM mapping file Tip Keep all files of a VM together in one folder for easier backup and management Chapter 7 VMware vSphere considerations 139 7 Select Compatibility Mode from the available options as shown in Figure 7 35 Physical The RDM will not be affected by snapshots similar to a virtual disk in mode independent persistent Tip
83. in working with the IBM System i platform and its predecessor models and eight years of experience in working with storage She has a Master s degree in Computer Science and a degree in Mathematics from the University of Ljubljana in Slovenia Copyright IBM Corp 2011 2013 All rights reserved xi Kai Jehnen is a Product Field Engineer for the DS8000 and has been working for the past seven years in this area He holds a degree in Information Technologies from the University of Applied Sciences in Koblenz His main focus is on solving critical customer situations in open systems environments He is also a VMware Certified Professional Steven Joseph is a Staff Software Engineer and Team Leader for the Product Engineering Tools Team in Tucson Arizona He has been with IBM for eight years starting as a Product Field Engineer for ESS and DS8000 Now as a developer he works on RAS features for the DS8000 family and data analysis tools for global Product Engineering teams for all Storage products He is a Certified Java Developer and has been developing AIX software for over 15 years He also holds multiple vendor certifications from HP Cisco and Sun Steven currently leads the development team for the ARK project producing tools for remote support and real time diagnostics for IBM Disk and Tape products Massimo Olivieri joined IBM in 1996 to work as Country Specialist for Tape Products He moved on to the Disks High End family as FE suppor
84. is set to be mirrored you will need to select them in pairs of equal capacity Selected disk units ede a A CS Ne Eligible Figure 11 21 Confirming disks to add to disk pool 232 IBM System Storage DS8000 Host Attachment and Interoperability 12 In the New Disk Pool Summary window Figure 11 22 click Finish to add disks to the disk pool New Disk Pool Summary The following shows how the configuration you specified will look after the disk units are added Ifthis is not correct click Back and make any necessary changes lfthis is correct click Finish to begin adding the disk units RAID 5 70 6 GB 21 07 A04 0 RAID 5 70 6 GB 2107 A04 0 RAID 5 70 6 GB 2107 A04 0 RAID 5 70 6 GB 2107 A04 0 RAID 5 70 6 GB 2107 A04 0 RAID 5 70 6 GB 2107 A04 0 RAID 5 70 6 GB 2107 A04 0 RAID 5 70 6 GB 2107 A04 0 Primary No Protected 282 3GB RAID 5 70 6 GB 2107 A04 0 RAID 5 70 6 GB 2107 A04 0 RAID 5 70 6 GB 2107 A04 0 RAID 5 70 6 GB 2107 A04 0 Figure 11 22 New Disk Pool Summary 13 Respond to any message windows that appear After you take action on any messages the New Disk Pool Status window shown in Figure 11 23 appears and shows progress New Disk Pool Status eal lia W Adding Disk Units oY Assigning Disk Pool Group Percent Complete 100 Figure 11 23 New Disk Pool Status Tip This step might take time depending on the number and size of the logical units that you add 14 When comple
85. logical volumes as backing devices at the VIOS level when intending to use two virtual I O servers and client AIX MPIO multipathing 66 IBM System Storage DS8000 Host Attachment and Interoperability When using a redundant dual VIOS setup with DS8000 devices as virtual SCSI backing devices and native AIX MPIO in the client partition follow the setup advice for the MPIO in the client partition in BM PowerVM Virtualization Introduction and Configuration SG24 7940 Figure 5 2 provides an example of MPIO in the client partition Client Partition hdisk devices in client algorithm failover ff ed em reserve_policy no_reseve Reet hcheck_mode nonactive hcheck_interval 60 hdisk devices on VIOS algorithm load_balance feserve_policy no_reserve hcheck_mode nonactive hcheck_interval 60 fscsi devices on VIOS dyntrk yes fc_erm_recov fast_fail __ Physical Resources L Virtual Resources Figure 5 2 Preferred MPIO attributes for DS8000 devices in a dual VIOS setup Using the default attributes for MPIO multipathing of the virtual SCSI disk devices in the AIX client partition as shown in Example 5 5 on page 65 set the hcheck_interval from 0 disabled to the preferred value of 60 seconds on the client partition in a dual VIOS setup as shown in Example 5 9 so that the path status is updated automatically The hcheck_interval attribute defines how often the health che
86. lowest tier The busiest workloads or the most mission critical data is said to be Aot relative to other application data that might be held on the same storage subsystem Typically put the hottest data on the fastest storage medium and colder data onto slower storage The DS8800 system can support the installation of all three storage types within its frames and then the ranks and extents can be provisioned accordingly And of course an application s needs can change over the lifetime of its data so a layout that makes sense today might not hold true tomorrow The Easy Tier software can constantly evaluate the temperature of the data as it is accessed and make changes to the storage between tiers and within a tier accordingly The hottest data can be provisioned onto the highest tier and other data can be moved as appropriate In addition the workloads on a tier can be balanced out making certain that storage units are not overused while other capacity exists Manual arrangement of data storage between the tiers is possible as well For more detailed information about Easy Tier and the performance characteristics of the DS8000 storage types see the Redpaper publication IBM System Storage DS8000 Easy Tier REDP 4667 02 IBM System Storage DS8000 Host Attachment and Interoperability Open systems considerations This chapter provides information about and general considerations for attaching IBM System Storage DS8000 serie
87. of attaching IBM System Storage DS8000 systems to host systems running Linux The following topics are covered gt Working with Linux and DS8000 Attaching to a basic host Resizing DS8000 volumes dynamically Using FlashCopy and remote replication targets Troubleshooting and monitoring vvvy Additionally this chapter provides information about the following hardware architectures that are supported for DS8000 attachment gt Intel x86 and x86_64 gt IBM Power Systems gt IBM System z Although older Linux versions can work with the DS8000 this chapter covers the most recent enterprise level distributions such as these products gt Novell SUSE Linux Enterprise Server 11 Service Pack 1 SLES11 SP1 gt Red Hat Enterprise Linux RHEL 5 Update 5 Red Hat Enterprise Linux 5U5 Red Hat Enterprise Linux 6 0 and Red Hat Enterprise 6 1 Copyright IBM Corp 2011 2013 All rights reserved 81 6 1 Working with Linux and DS8000 Linux is an open source UNIX like operating system that uses the Linux kernel This chapter presents information for working with the DS8000 and the Linux operating system The Linux kernel along with the tools and software needed to run an operating system are maintained by a loosely organized community of thousands of mostly volunteer programmers 6 1 1 How Linux differs from other operating systems Linux differs from the other proprietary operating systems in many ways such as the follow
88. panel on the right Here you can view and change the Frame Size Connection Mode Date Rate and many other options This panel also allows the user to reset all the fields to their default values To make NVRAM changes permanent be sure to perform all three of the following steps 1 Make the change on each channel as needed 2 Click the Commit button 3 Restart the Apple server See the IBM System Storage DS8000 Host Systems Attachment Guide GC27 2298 02 for any special settings that might be needed 8 5 Creating a file system on Apple Mac OS X You can create a file system using either the GUI or the CLI of the Mac OS X Server This section describes the use of both methods 156 IBM System Storage DS8000 Host Attachment and Interoperability 8 5 1 Using the GUI Disk Utility The Disk Utility is a standard part of Mac OS X It can be found under Utilities or by clicking initialize when a pop up window appears for a newly created LUN after detection From the left pane locate the disk you want to use and click Partition In the Volume Scheme select the number of partitions Select the volume format and provide a volume name After clicking Apply you need to click Partition to start the format 8 5 2 Using the CLI diskutil The disktutil command is used to partition repair or erase the volume Use this command to list the attached disks to the host system as shown in Example 8 2 Example 8 2 Using diskutil macintosh 4
89. ports is in zone with multiple DS8000 ports IBM i will still use only one of them for I O traffic If multiple IBM i ports are in the same zone with multiple DS8000 ports each IBM i port might use the same DS8000 port for the IO traffic Thus one of the DS8000 ports becomes overloaded and the others are idle which causes unbalanced usage of DS8000 ports and consequently impact on DS8000 performance DS8000 host adapters can be shared between System i and other platforms gt For performance reasons it is best to limit the number of IBM i adapters or ports connected to one Host attachment HA card in DS8000 Based on the available measurements and experiences consider planning the following maximal number of IBM i ports for one port of HA card in DS8000 With lIOP based adapters e Four adapters per HA card port in DS8100 DS8300 and DS8700 e Up to 12 adapters per HA card in DS8800 With 4 GB IOP less adapters e Two to four ports per HA card in DS8100 DS8300 and DS8700 e Six to ten ports per HA card in DS8800 With 8 GB IOP less adapters e Two ports per HA card in DS8100 DS8300 and DS8700 e Four to eight ports per HA card in DS8800 For a current list of switches supported under IBM i see the web pages System Storage Interoperation Center and IBM i Storage Solutions as listed in 11 1 4 Useful websites on page 204 11 5 Using multipath With IBM i multipath is part of the base operating sy
90. possible to assign an MIH value to an alias address Alias devices are not known externally and are only known and accessible by IOS If an external method is used to attempt to set the MIH on an alias device address an IOS090l1 message is generated For example the following message is generated for each attempt to set the MIH on an alias device IOSO090 alias device number S AN INVALID DEVICE Tip When setting MIH times in the IECIOSxx member of SYS1 PARMLIB do not use device ranges that include alias device numbers 12 5 3 HyperPAV z OS support and implementation The DS8000 series users can benefit from enhancements to Parallel Access Volume PAV with support for HyperPAV HyperPAV provides the ability to use an alias address to access any base on the same logical control unit image per I O base Benefits of HyperPAV HyperPAV was designed to offer the following capabilities gt Provide an even more efficient Parallel Access Volumes PAV function gt Help clients who implement larger volumes to scale I O rates without the need for additional PAV alias definitions gt Exploit the FICON architecture to reduce impact improve addressing efficiencies and provide storage capacity and performance improvements More dynamic assignment of PAV aliases improves efficiency The number of PAV aliases needed can be reduced taking fewer from the 64 K device limitation and leaving more storage for capacity use gt Enabl
91. ranks An extent pool is visible to both processor complexes in the DS8000 but it is directly managed by only one of them You must define a minimum of two extent pools with one extent pool created for each processor complex to fully use the resources The current approach considering the availability of Storage Pool Striping and Easy Tier for the DS8000 is to create a few DS8000 only extent pools for instance two Then use either DS8000 storage pool striping or automated Easy Tier rebalancing to help prevent from overloading individual ranks You need only one MDisk volume size with this approach because plenty of space is available in each large DS8000 extent pool Often clients choose 2 TiB 2048 GiB MDisks with this approach Create many 2 TiB volumes in each extent pool until the DS8000 extent pool is full and provide these MDisks to the SAN Volume Controller to build the storage pools You can find a more detailed discussion in the IBM Redbooks publication DS8800 Performance Monitoring and Tuning SG24 8013 IBM System Storage DS8000 Host Attachment and Interoperability Related publications The publications listed in this section are considered particularly suitable for a more detailed discussion of the topics covered in this book IBM Redbooks publications The following IBM Redbooks publications provide additional information about the topic in this document Some publications referenced in this list might be available i
92. role of a remote copy services relationship Metro Mirror Management improvements The new Metro Mirror Management MGM improvements require APAR OA37632 Tivoli Storage Productivity Center for Replication for System z provides improved management capability for Metro Mirror sessions containing system volumes Enablement of HyperSwap is no longer required This feature also helps ensure that data consistency is still maintained even if Tivoli Storage Productivity Center for Replication for System z is not active at the time of a disaster 278 IBM System Storage DS8000 Host Attachment and Interoperability Improved management and alerting The TPC R for System z provides the ability to set a warning and a severe threshold on Recovery Point Objective RPO metrics for a Global Mirror session If the RPO surpasses the warning threshold an SNMP event will be thrown and a message issued for the Tivoli Storage Productivity Center for Replication for System z session indicating that the RPO has passed the warning point and therefore might soon pass the severe threshold If the RPO surpasses the severe threshold an SNMP event will be activated a message issued and the session should indicate a severe status until the RPO again drops below the threshold Additionally Tivoli Storage Productivity Center for Replication for System z allows the user to export Global Mirror history to a CSV file for RPO and Out of Sync analysis This feature
93. run the fsadm command successfully you need the appropriate license If you do not have the license installed you will see the following message UX vxfs fsadm ERROR V 3 25255 fsadm You don t have a license to run this program The number following the b option can be calculated as follows The new size of the logical volume equals 4 600 physical extends PEs The PE size equals 4 MB Therefore in total the new size equals to 4 MB times 4 600 PEs which equals 18 400 MB or 18 400 000 bytes 10 8 2 Working with large LUNs Large LUNs are multiple terabytes in size The information presented up to this point cannot be applied when working with large LUNs A new set of parameters and options are required The rest of this chapter provides information about working with large LUNs Chapter 10 HP UX considerations 197 Example 10 18 depicts the following situations gt The lvmadm t command demonstrates attaching support for the Version 2 1 volume group gt The properties shown in bold highlight the differences between volume group versions 2 1 2 0 and 1 0 Example 10 18 Showing the properties of separate volume group versions Tvmadm t LVM Limits VG Version Max VG Size Tbytes Max LV Size Tbytes Max PV Size Tbytes Max VGs Max LVs Max PVs Max Mirrors Max Stripes Max Stripe Size Kbytes Max LXs per LV Max PXs per PV Max Extent Size Mbytes VG Version Max VG S
94. show up as configured Solaris 10 In Solaris 10 the cfgadm command is only needed if the LUNs are not visible when issuing the format command after having mapped the LUNs to the Solaris host 164 IBM System Storage DS8000 Host Attachment and Interoperability Solaris 8 and 9 To configure MPxIO in Solaris 8 and 9 enable it by editing the kernel drv scsi_vhci conf file Find the mpxio disable parameter and change the setting to no Example 9 7 Example 9 7 Changes to kernel drv scsi_vhci conf mpxio disable no Solaris 10 To enable MPxIO in Solaris 10 complete these steps 1 Execute the stmsboot e command 2 Add the following stanza to supply the vendor identification VID and product identification PID information to MPxIO in the kernel drv scsi_vhci conf file device type scsi options list TBM 2107900 symmetric option symmetric option 0x1000000 Attention The vendor string must be exactly 8 bytes so you must type IBM followed by five spaces 3 Restart the system with the reboot command After the restart MPxIO is ready to be used For example the LUNs can now be prepared with the format command and the partitions and filesystems can be created For more information about MPxIO see the following websites gt See the Oracle Solaris I O Multipathing page at this website http download oracle com docs cd E19680 01 821 1253 intro 46 index htm gt See the Oracle Wiki home page for MPxIO at
95. so no utility is needed 9 6 1 Displaying the boot code In Example 9 21 you can see that the boot code installed is FCode BIOS Version 3 10a3 If the boot code is not installed on your HBAs download it from the Emulex website Issue the fcinfo command to see the boot code installed and general HBA information Example 9 21 Displaying the boot code bash 3 00 fcinfo hba port HBA Port WWN 10000000c95f3592 OS Device Name dev cfg c4 Manufacturer Emulex Model LP10000 Firmware Version 1 92a1 T2D1 92A1 FCode BIOS Version 3 10a3 Serial Number MS65035391 Driver Name emlxs Driver Version 2 31h 2008 06 16 08 54 Type N port State online Supported Speeds 1Gb 2Gb Current Speed 2Gb Node WWN 20000000c95f3592 HBA Port WWN 10000000c95f35dc OS Device Name dev cfg c3 Manufacturer Emulex Model LP10000 Firmware Version 1 92a1 T2D1 92A1 FCode BIOS Version 3 10a3 Serial Number MS65035465 Driver Name emlxs Driver Version 2 31h 2008 06 16 08 54 Type N port State online Supported Speeds 1Gb 2Gb Current Speed 2Gb Node WWN 20000000c95f35dc When the Fibre Channel adapters are prepared with the boot code as described they can be used to install a new instance of Solaris on a LUN that resides in the SAN Chapter 9 Solaris considerations 173 9 6 2 Booting off a DS8000 LUN with Solaris Example 9 22 shows a Solaris system that was installed from a CDROM on a DS8000 LUN with file systems residing on it Solaris can
96. straddle where the cylinder managed area begins gt Exact space will be obtained if allocated on a track managed area exactly how it was before EAV implementation gt If the requested space is not available from the preferred managed space as determined by BPV the system can allocate the space from both cylinder managed and track managed spaces gt Because VSAM striped data sets require that all the stripes be the same size the system will attempt to process exact requested space using a combination of cylinder managed and track managed spaces VSAM control area considerations For VSAM control area CA keep in mind the following considerations gt VSAM data sets created in z OS V1 10 for EAV and non EAV might have separate CA sizes from what was received in prior releases The reason for it is that a CA must be compatible with an MCU for a cylinder managed area gt VSAM data sets allocated with compatible CAs on a non EAV are eligible to be extended to additional volumes that support cylinder managed space VSAM data sets physically copied from a non EAV to an EAV might have an incompatible CA and might not be eligible for EAV It means that extents for additional space do not use cylinder managed space Migration consideration This consideration is for a migration to z OS V1 10 12 4 3 EAV migration considerations 260 For EAV migration keep in mind the following considerations gt Assistance for migration will
97. that the OS loader can access this disk FC HBAs provide an extension to the system firmware for this purpose In many cases it must be specifically activated On x86 systems this location is called the master boot record MBR Tip For Linux under z VM the OS loader is not part of the firmware but the z VM ip program The boot loader The purpose of the boot loader is to start the operating system kernel To do this it must recognize the physical location of the kernel image on the system disk read it in unpack it if it is compressed and start it All of it is done using the basic I O routines provided by the firmware The boot loader can also pass configuration options and the location of the InitRAMFS to the kernel The following most common Linux boot loaders are gt Grand unified boot loader GRUB for x86 systems gt zipl boot loader for System z gt yaboot for Power Systems The kernel and the initRAMFS After the kernel is unpacked and running it takes control over the system hardware It starts and sets up memory management interrupt handling and the built in hardware drivers for the hardware that is common on all systems such as the memory management unit clock and so on It reads and unpacks the initRAMFS image using the same basic I O routines The initRAMFS contains additional drivers and programs that are needed to set up the Linux file system tree or root file system To be able to boot from a SAN attache
98. the timeout_policy to retry_path first otherwise the changing device algorithm command will fail gt Incase the device s algorithm is set to fail_over with SDDPCM version before 2 6 3 0 and the user wants to upgrade SDDPCM to v2 6 3 x then additional steps are required to properly configure these devices For detailed information see the following SDDPCM flash http www 01 ibm com support docview wss rs 540 amp context ST52G7 amp dc D600 amp ui d ssg1 1004072 amp 1oc en_US amp cs utf 8 amp lang en 5 3 3 max_xfer_size attribute There are some cases where there is too much I O and the adapter cannot handle all of the requests at once It is common to find errors indicating that the host adapter is unable to activate an I O request on the first attempt The most likely cause of these errors is that the host is low on resources To reduce the incidence of these errors follow these examples to increase the resources by modifying the maximum transfer size attribute for the adapter The default max_xfer size value is 0x1000000 It is possible to check the values as shown in Example 5 13 Example 5 13 Checking the max_xfer_size value lsattr El fcsO grep max_xfer_size max_xfer_size 0x100000 Maximum Transfer Size True Tip To view the range of allowed values for the attribute type lsattr RI adapter_name a max_xfer_size Example 5 14 shows how to increase the size of the setting Example 5 14 Changing
99. the max_xfer_size parameter chdev 1 fcsO P a max_xfer_size 0x400000 fcs0 changed Tip After changing the setting it will then require a host reboot to take effect 5 3 4 Storage unit with multiple IBM Power Systems hosts running AIX On a host with non MPIO SDD installed a reserve bit is placed on all disks that are part of the volume group when that volume group is varied online But with MPIO SDDPCM the no_reserve policy in AIX is changed to no_reserve It means that if the rootvg volume is shared with another host the data might be accidentally damaged Chapter 5 AIX considerations 71 5 4 Multipathing with AIX Information in this section provides configurations and settings for AIX hosts with DS8000 The DS8000 supports two methods of attaching AIX hosts gt Subsystem Device Driver SDD gt AIX multipath I O MPIO with a DS8000 specific Path Control Module SDDPCM SDD and SDDPCM cannot be installed on the same AIX server On an AIX server you need to have installed either the SDD or SDDPCM file sets for the DS8000 volumes gt Filesets for SDD devices fcp disk ibm rte devices sdd 61 rte here 61stands for AIX6 1x gt Filesets for SDDPCM devices fcp disk ibm mpio rte devices sddpcm 71 rte here 71 stands forAlX7 1 The SDD or SDDPCM file sets require separate DS8000 host attachment file sets ibm fcp disk ibm installed as a prerequisite You can only have SDD or SDDPCM file sets installed
100. this website http wikis sun com dosearchsite action queryString MPx10 gt See the Oracle Documentation page at this website http www oracle com technetwork indexes documentation index htm 9 4 3 Working with VERITAS Volume Manager dynamic multipathing Before using VERITAS Volume Manager VxVM DMP a part of the VERITAS Storage Foundation suite make sure you have downloaded and installed the latest maintenance pack You also need to download and install the Array Support Library ASL for the DS8000 To download ASL select the respective product from the following website https sort symantec com as Upon selecting an item such as Patches you are redirected to the Veritas Operations site Using VxVM DMP During device discovery the vxconfigd daemon compares the serial numbers of the separate devices If two devices have the same serial number then they are the same LUN and DMP combines the paths Listing the contents of the dev vx rdmp directory provides only one set of devices Chapter 9 Solaris considerations 165 The vxdisk path command also demonstrates DMP s path suppression capabilities In Example 9 8 you can see that device c7t2d0s2 is suppressed and is only shown as a subpath of c6t1d0s2 Example 9 8 vxdisk path command output vxdisk path SUBPATH DANAME DMNAME GROUP STATE c6t1d0s2 c6t1d0s2 Ethan0l Ethan ENABLED c7t2d0s2 c6t1d0s2 Ethan0l Ethan ENABLED c6tld1s2 c7t2d1s2 Ethan02 Ethan ENABLED c7t2d1s2
101. transition 72 IBM System Storage DS8000 Host Attachment and Interoperability Subsystem Device Driver for AIX does not use native AIX MPIO Every path to each disk or volume is represented by a logical hdisk device in a system On top of the paths the new logical device vpath is created which must be used as a logical device instead of hdisk as shown in Example 5 15 Example 5 15 Isvpcfg command Isvpcfg vpathO Avail pv vg 2200 75TV1812200 vpathl Avail pv vg 2201 75TV1812201 vpath2 Avail pv vg_2300 75TV1812300 vpath3 Avail pv vg_2301 75TV1812301 hdisk2 Avail hdisk6 Avail hdisk10 Avail hdisk14 Avail hdisk3 Avail hdisk7 Avail hdisk11 Avail hdisk15 Avail hdisk4 Avail hdisk8 Avail hdisk12 Avail hdisk16 Avail hdisk5 Avail hdisk9 Avail hdisk13 Avail hdisk17 Avail Tip To create a volume group using SDD always use mkvg4vp instead of the AIX default mkvg command The vpath devices have Port VLAN ID PVID but the disks do not As shown in Example 5 16 paths are identified as an IBM FC 2107 device while vpath devices are identified as Data Path Optimizer Pseudo Device Driver Example 5 16 List disk devices on an AIX system using the Isdev command Isdev Cc disk hdisk2 Availabl hdisk3 Availabl hdisk4 Availabl hdisk5 Availabl hdisk6 Availabl hdisk7 Availabl hdisk8 Availabl hdisk9 Availabl hdisk10 Availabl hdisk11 Availabl hdisk12 Availabl hdisk13 Availabl hdisk14 Ava
102. workload is not meeting its goal due to IOS queue IOSQ time WLM will attempt to find an alias device that can be reallocated to help this workload achieve its goal gt HyperPAV Dynamic PAV requires the WLM to monitor the workload and goals It takes some time until the WLM detects an I O bottleneck Then the WLM must coordinate the reassignment of alias addresses within the sysplex and the DS8000 All of this takes time and if the workload is fluctuating or has a burst character the job that caused the overload of one volume could have ended before the WLM had reacted In these cases the OSQ time was not eliminated completely With HyperPAV the WLM is no longer involved in managing alias addresses For each I O an alias address can be picked from a pool of alias addresses within the same LCU This capability also allows different HyperPAV hosts to use one alias to access different bases which reduces the number of alias addresses required to support a set of bases in an IBM System z environment with no latency in assigning an alias to a base This functionality is also designed to enable applications to achieve better performance than is possible with the original PAV feature alone while also using the same or fewer operating system resources You can find more information regarding dynamic PAV at the following website http www ibm com systems z os zos features wlm Again consider HyperPAV over dynamic PAV management which allows
103. written by vendors but the MPIO driver package by itself does not form a complete solution This joint solution helps storage vendors provide device specific solutions that are tightly integrated with the Windows operating system MPIO drivers MPIO is not shipped with the Windows operating system Storage vendors must pack the MPIO drivers with their own DSM IBM Subsystem Device Driver Device Specific Module SDDDS M is the IBM multipath I O solution based on Microsoft MPIO technology It is a device specific module specifically designed to support IBM storage devices on Windows 2003 servers MPIO helps integrate a multipath storage solution with the operating system and provides the ability to use multipathing in the SAN infrastructure during the boot process for SAN boot hosts 22 IBM System Storage DS8000 Host Attachment and Interoperability 3 5 Partition alignment Partition alignment is an important performance process that is often overlooked Windows 2003 can become considerably degraded when working with multiple disks if the partition alignment is not set properly Windows 2008 attempts to set it up by default but Windows 2003 does not cover this concept This process was noted to be required by Windows servers below 2008 Setting up the proper disk alignment can create a proper Master Boot Record MBR which can then allocate the proper unit size per partition created See the following links for additional information about di
104. 0 1 102 Trks Cyls 15 Free DSCBS 7 497 Free Tracks 16 695 Tracks 150 Size 16 530 1 102 Extents 1 F1 Help F2 Split F3 Exit F9 Swap F12 Cancel Example 12 3 shows a typical VTOC summary for a 3390 Model A EAV volume It is divided into two parts the new Tracks managed and the Total space Example 12 3 VTOC information for 3390 A 1TB VTOC Summary Information Volume RE 7BF Command gt Unit 3390 Free Space Chapter 12 IBM System z considerations 257 VTOC Data Total Tracks Cyls Tracks 150 Size 17 729 925 1 181 995 Used 1 Largest 16 747 290 1 116 486 Free DSCBS 7 497 Free Extents 2 Volume Data Track Managed Tracks Cyls Tracks 17 730 090 Size a os 982 635 65 509 Used 0 Largest 982 635 65 509 Trks Cyls 15 Free F1l Help F2 Split F3 Exit F9 Swap F12 Cancel The DEVSERV command can also be used to identify an EAV volume Example 12 4 shows the new value for the DTYPE field by running the DEVSERV PATHS command Example 12 4 DEVSERV command for 3390 mod3 and 3390 mod A 1TB ds p be 2 IEE459I 13 00 47 DEVSERV PATHS 646 UNIT DTYPE M CNT VOLSER CHPID PATH STATUS RTYPE SSID CFW TC DFW PIN DC STATE CCA DDC CYL CU TYPE 077BE 33903 0 000 RE77BE 80 81 85 86 82 83 84 87 2107 7701 Y YY YY N SIMPLEX BE BE 1113 2107 077BF 3390A 0 000 RE77BF 80
105. 000 Host Systems Attachment Guide GC27 2298 02 at this website http www 304 ibm com support docview wss uid ssg1S7001161 amp aid 1 Sun operating system Oracle has acquired Sun For this reason the operating system that was known as Sun Solaris was officially renamed as Oracle Solaris Likewise the Sun Cluster also known as the Solaris Cluster is now called the Oracle Solaris Cluster Before Solaris can access LUNs on a DS8000 some prerequisites need to be met gt The Solaris server is equipped with a Fibre Channel host bus adapter HBAs For redundancy it is best to have two HBAs gt The Solaris server is physically connected with Fibre Channel cabling to a SAN to which the DS8000 is also attached gt SAN zoning is configured to allow the server HBAs to communicate with the DS8000 host adapters When this preparation work is complete you can configure the LUNs in the DS8000 to be accessed by the Solaris server It is described in the following chapters 9 2 Locating the WWPNs of your HBAs For a server to access LUNs configured in a DS8000 storage system these LUNs need to be mapped to the server HBA s WWPNs An object in the DS8000 is created for each WWPN called hostconnection One or more hostconnections can be mapped to one or more LUNs in the DS8000 by adding those LUNs to an object called a volume group The mapping in the DS8000 is managed by connecting a hostconnection to a volume group containing the LUNs I
106. 000 natively specify the WWPN of the port in IBM i When connecting with VIOS NPIV specify the WWPN of the port in virtual Fibre Channel adapter in IBM i When connecting DS8000 natively with IOP less or IOP based adapters set up the ports in DS8000 as follows gt With point to point connection define the ports as Fibre Channel _AL An exception is to use SCSI Fibre Channel protocol SCSI FCP for ports that connect the load source LUN with lOP based adapters through point to point connection gt If the connection is through switches define the ports as SCSI FCP IBM System Storage DS8000 Host Attachment and Interoperability Figure 11 6 shows how to define the host connection for IBM i attached with SAN switches Create New Host Connection Define Host Ports Define one or more host ports that you will use to map hosts to a volume group in the next step After you add table will be mapped to the same volume group when you create the connection Host Connection Nickname Port Type Host 1 Fibre Channel Pointto Point Switched FeSf a Host Type IBM pSeres RS 6000 and RS 6000 SP Servers AlX ipSeres v Enter the 16 digit WWPN manually or select it from the list and then add it to the table Host WWPN 10000000038315 E IFE Select WWPN Nickname Figure 11 6 Create a host connection for IBM i in DS8000 GUI The following example shows how to define a host connection for IBM i
107. 04 Unformatted capacity 18432 000 MBytes Write Cache Enabled Read Cache Enabled Minimum prefetch 0x0 Maximum prefetch 0x16 Device Path s Type Disk device dev rdsk c2t50050763030CC1A5d0s2 devices pci 9 600000 SUNW q1c 1 fp 0 0 ssd w50050763030cc1a5 0 c raw dev rdsk c3t50050763031CC1A5d0s2 devices pci 9 600000 SUNW q1c 2 fp 0 0 ssd w50050763031cc1a5 0 c raw The disk now has a capacity of 18 GB To use the additional capacity issue the vxdisk resize command as shown in Example 9 17 Example 9 17 VxVM commands after volume expansion bash 3 00 vxdisk resize IBM_DS8x001_0 bash 3 00 vxdisk list IBM_DS8x001_0 Device devicetag type hostid disk group info flags pubpaths guid udid site version josize public private update ssb headers configs logs IBM_DS8x001_0 IBM_DS8x001_0 auto v880 name IBM_DS8x001_0 id 1194446100 17 v880 name 20781_ dg id 1194447491 20 v880 format cdsdisk privoffset 256 pubslice 2 privslice 2 online ready private autoconfig autoimport imported block dev vx dmp IBM_DS8x001_0s2 char dev vx rdmp IBM_DS8x001_0s2 fbdbfe12 1dd1 11b2 af7c 0003ba43fdc1 1BM 5F2107 5F7520781 5F6005076303FFC1A50000000000004704 3 1 min 512 Bytes max 2048 blocks slice 2 offset 65792 len 37677568 disk_offset 0 slice 2 offset 256 len 65536 disk_offset 0 time 1194473744 seqno 0 16 actual_seqno 0 0 0 240 count 1 len 48144 count 1 len 7296 Defined regio
108. 0C9808619 Example 5 1 shows two physical fcs0 fcs1 Fibre Channel adapters in the AIX partition Alternatively if you have SDD or SDDPCM already installed on your machine see 5 4 Multipathing with AIX on page 72 for details you can use either the datapath or the pempath command to get the list of WWPNs Both commands use the same syntax Example 5 2 shows a list of WWPNs after using the pcmpath command Example 5 2 SDD WWPN query pcempath query wwpn Adapter Name PortWWN fscsi0 10000000C9808618 fscsil 10000000C9808619 5 1 1 Assigning volumes Complete the following steps on the DS8000 storage system to directly assign volumes to an AIX partition using Fibre Channel adapters 1 Create a DS8000 volume group for the volumes accessed by the AIX server using the mkvolgrp command with hosttype pSeries 2 Create DS8000 host connections for the Fibre Channel adapters in the AIX server using the mkhostconnect command with hosttype pSeries and an unused port group for example port group 45 to group multiple host connections of the same host or hosts sharing the same volumes It simplifies connection management later Also you can assign a volume group to a host connection definition at this time or do it later using the managehostconnect command 3 Create and assign DS8000 volumes to DS8000 volume group using the mkfbvol command or later using chvolgrp command Example 5 3 shows these examples extent pools P8 and P9 as sto
109. 13 cylinders boundary sizes 56 763 51 x 1113 to 1 182 006 1062 x 1113 cylinders The size of an existing Model 3 9 A volume can be increased to its maximum supported size using dynamic volume expansion It can be done with the DS CLI command as shown in Example 12 1 or by using the web GUI Example 12 1 Dynamically expand CKD volume dscli gt chckdvol cap 1182006 captype cyl quiet 90f1 Date Time 8 September 2010 1 54 02 PM IBM DSCLI Version 6 6 0 284 DS IBM 2107 75TV181 CMUC00022I chckdvol CKD Volume 90f1 successfully modified DVE Dynamic Volume Expansion can be done as the volume remains online with the host system and is configured in the following increments gt Modi 1 113 cylinders on volumes gt than 65 520 cylinders gt One cylinder on volumes with lt 65 520 cylinders A VTOC refresh through ICKDSF is a preferred practice use ICKDSF REFORMAT REFVTOC to rebuild a VTOC index to be full compatible with the new cylinder addressing even if the volume will remain online showing the new number of cylinders using the DEVSERV command as shown in Example 12 4 on page 258 256 IBM System Storage DS8000 Host Attachment and Interoperability Volume initialization Now with the new function QUICK INIT volume initialization is performed dynamically after volume expansion with no delay for initialization before going online to the host The VTOC allocation method for an EAV volume is changed compared to the V
110. 16 active ready running 4 0 1 4 sdu 65 64 active ready running 4 0 0 4 sdt 65 48 active ready running 5 As shown in Example 6 53 mount the target volume to a separate mount point using a device node that is created from the unique identifier of the volume Example 6 53 Mounting the target volume x36501ab9 mnt mount dev mapper 3600507630af fc29 f00000000000047b1 mnt itso_fc x36501ab9 mnt mount dev mapper 3600507630af fc29f0000000000004791 on mnt itso_4791 type ext3 rw dev mapper 3600507630af fc29f00000000000047b1 on mnt itso fc type ext3 rw Now you can access both the original volume and the point in time copy through their respective mount points This demonstration used an ext3 file system on a DM Multipath device which is replicated with FlashCopy Attention The device manager udev also creates device nodes that relate to the file system unique identifier UUID or label These IDs are stored in the data area of the volume and are identical on both the source and target Such device nodes are ambiguous if the source and the target are mapped to the host at the same time Using them in this situation can result in data loss 6 4 2 Using a file system residing in a logical volume managed by LVM The Linux LVM uses metadata that is written to the data area of the disk device to identify and address its objects If you want to access a set of replicated volumes that are under LVM control this metadata must b
111. 2 H3 H4 ABCD ABCD ABCD ABCD ABCD ABCD ABCD ABCD ABCD ABCD ABCD ABCD ABCD ABCD ABCD ABCD BAY 5 B5 BAY 6 B6 BAY 7 B7 BAY 8 B8 H1 H2 H3 H4 H1 H2 H3 H4 H1 H2 H3 H4 H1 H2 H3 H4 ABCD ABCD ABCD ABCD ABCD ABCD ABCD ABCD ABCD ABCD ABCD ABCD ABCD ABCD ABCD ABCD 81 Disk rene Y Y V1 Disk3 Y 000 Y eee eee 61 Disk4 rene Y Y online open y alternate path online open online closed o alternate path online close offline n alternate path offline path not configured path information not available path down 2105 devices essid has 5 digits while 1750 2107 device s essid has 7 digits Chapter 3 Windows considerations 25 The datapath query essmap command provides additional information about LUNs and I O port numbers as shown in Example 2 6 Example 3 6 datapath query essmap C Program Files IBM SDDDSM gt datapath query essmap Disk Path P Location LUN SN Type Size LSS Vol Rank C A S Connection Port RaidMode Disk2 Path0 Portl BusO 75207814703 IBM 2107900 12 0GB 47 03 0000 2c Y R1 B2 H4 ZD 143 RAIDS Disk2 Pathl Port2 BusO 75207814703 IBM 2107900 12 0GB 47 03 0000 2c Y R1 B4 H4 ZD 343 RAIDS Disk3 Path0 Portl BusO 75ABTV14703 IBM 2107900 12 0GB 47 03 0000 Ob Y R1 B1 H1 ZA 0 RAID5 Disk3 Path1 Port2 Bus0 7
112. 27 4055 gt IBM Tivoli Storage Productivity Center for Replication for System z V5 1 Installation and Configuration Guide SC27 4053 gt IBM Tivoli Storage Productivity Center for Replication for System z V5 1 User s Guide C27 4054 Chapter 12 IBM System z considerations 279 12 10 Full Disk Encryption FDE Encryption disks must be managed by a key server usually a couple of dedicated TKLM servers that provide full access to the DS8000 When encryption is enabled on a DS8000 system the disk is locked when a rank is created assigning an access credential It occurs before data is stored To access data the DS8000 system must provide the access credential each time a locked disk is powered on If a rank is deleted the disk generates a new encryption key which replaces the existing encryption key and crypto erases the disk The DS8000 generates the access credential for a locked IBM FDE drive using a data key that it obtains from the Tivoli Key Lifecycle Manager After this step completes the DS8000 can write and read data on the disk with the disk encrypting data on writes and decrypting data on reads Without access to the Tivoli Key Lifecycle Manager the DS8000 system does not know the disk access credential and cannot read or write data on the disk When an encryption group is configured on the DS8000 system the system obtains a data key from the Tivoli Key Lifecycle Manager To prevent key server encryption deadlock you mig
113. 4 active ready running Use reconfigure as shown in Example 6 28 to configure DM Multipath to update the topology after scanning the paths and configuration files You can also use it to add new multipath devices after adding new DS8000 volumes For more information see Adding and removing DS8000 volumes dynamically on Linux on page 106 Example 6 28 Reconfiguring DM Multipath multipathd gt reconfigure ok Exit The multipathd k command prompt of SLES11 SP1 supports the quit and exit commands to terminate Press CTRL D to terminate or quit the session in Red Hat Enterprise Linux 5U5 Fast path You can also issue commands in a one shot mode by enclosing them in double quotes and typing them directly without space behind the multipath k command for example multipathd k show paths Print Although the multipath 1 and multipath 11 commands can be used to print the current DM Multipath configuration use the multipathd k interface The multipath tool will be removed from DM Multipath and all further development and improvements go into multipathd Accessing DM Multipath devices in SLES 11 The device nodes used to access DM Multipath devices are created by udev in the dev mapper directory If you do not change any settings SLES 11 uses the unique identifier of a volume as device name as shown in Example 6 29 102 IBM System Storage DS8000 Host Attachment and Interoperability Example 6 29 Multipath devices in SLES
114. 5ABTV14703 IBM 2107900 12 0GB 47 03 0000 Ob Y R1 B2 H3 ZA 130 RAID5 Disk4 Path0 Port1 BusO 75034614703 IBM 2107900 12 0GB 47 03 0000 Oe Y R1 B2 H4 ZD 143 RAIDS Disk4 Pathl Port2 BusO 75034614703 IBM 2107900 12 0GB 47 03 0000 Oe Y R1 B4 H2 ZD 313 RAIDS 3 6 3 Windows 2008 and SDDDSM Windows 2008 only supports SDDDSM The datapath query commands in Windows 2008 have not been changed in comparison to Windows 2003 The Server Manager window has changed slightly on Windows 2008 servers as shown in Figure 3 6 Ai REE Fie Action View Help T E a EN Server Manager WIN 3QLKU9EFUUH Disk Management Volume List Graphical View Actions 3 Roles aia Vole e e Fie System status Disk Managem m Diagnostics amp New Volume E Simple Basic NTFS Healthy Primary Partition More Actions gt F Event Viewer Volume G Simple Basic NTFS Healthy Primary Partition F g Reliability and Performance amp New Volume H Simple Basic NTFS Healthy Primary Partition g Device Manager Spanned Dynamic NTFS Healthy at Configuration GNew Volume J Simple Basic NTFS Healthy Primary Partition Task Scheduler G New Volume K Mirror Dynamic NTFS Healthy Q Windows Firewall with Advanced Security GaaSystem Drive D Simple Basic NTFS Healthy Active Primary Partition Gh Services GaxSystem Boot C Simple Basic NTFS Healthy System Boot Page File Active Crash Dump WMI
115. 63 0a 00 42 9F 0 Active ymhba3 C0 T0 L0 50 05 07 63 0a ff c2 9f 50 05 07 63 0a 00 42 9F 0 Active Ez vmhba3 C0 T1 L0 50 05 07 63 0a ff c2 9f 50 05 07 63 0a 40 42 9F 0 Active vmhba2 C0 T0 LO 50 05 07 63 0a ff c2 9f 50 05 07 63 0a 00 42 9F 0 Active x Figure 7 44 Change path selection plug in 3 The status field for the different paths should all change to Active I O Chapter 7 VMware vSphere considerations 145 7 4 Storage vMotion vSphere has a built in feature called Storage vMotion It provides the ability to live migrate virtual machines from one datastore to another without interruption Storage vMotion works independently from the underlying storage and is initiated by the GUI This section presents the following usage scenarios gt Migrating VMs from one storage device to another for example technology change gt Evacuating a storage device for maintenance gt Load balancing datastore usage 7 4 1 Steps to migrate a VM from one datastore to another To live migrate a VM onto another datastore perform the following steps 1 Select the VM from the list 2 Right click the VM and select Migrate as shown in Figure 7 45 Name State St E Win VM1 Power Guest gt Snapshot i dtt Settings E Migrate 2a Clane Figure 7 45 Start VM migration E Open Console Si 3 In the wizard select Change Datastore and click Next Figure 7 46 Chan
116. 64 DMP015 1 50 210529F 2107 A04 70564 DMP014 1 50 210429F 2107 A04 70564 DMP013 1 50 210329F 2107 A04 70564 DMP012 F3 Exit F5 Refresh Fll Display disk configuration capacity F12 Cancel Figure 11 11 Specifying the ASPs to add units Chapter 11 IBM i considerations 227 7 Inthe Confirm Add Units panel Figure 11 12 press Enter to continue if everything is correct Depending on the number of units you are adding this step can take time When it completes open your disk configuration to verify the capacity and data protection Confirm Add Units Add will take several minutes for each unit The system will have the displayed protection after the unit s are added Press Enter to confirm your choice for Add units Press F9 Capacity Information to display the resulting capacity Press Fl0 Confirm Add and Balance data on units Press F12 Cancel to return and change your choice Serial Resource HotSpare ASP Unit Number Type Model Name Protection protection 1 Unprotected 1 50 210729F 2107 A04 ccDMP016 RAID 5 N 2 50 200029F 2107 A04 ccDMPOO1 RAID 5 N 3 50 200129F 2107 A04 ccDMPO02 RAID 5 N 4 50 200229F 2107 A04 ccDMP003 RAID 5 N 5 50 200329F 2107 A04 ccDMP004 RAID 5 N 6 50 200429F 2107 A04 ccDMP005 RAID 5 N F9 Resulting Capacity F10 Add and Balance Fll Display Encryption Status F12 Cancel Figure 11 12 Confirming add units 11 6 10 Adding volumes to an independent auxiliary storage pool IASPs can be switc
117. 8 08 rx6600 1 vmunix DIAGNOSTIC SYSTEM WARNING May 30 12 38 08 rx6600 1 vmunix The diagnostic logging facility has started receiving excessive May 30 12 38 08 rx6600 1 vmunix errors from the I O subsystem 1 0 error entries will be lost 190 IBM System Storage DS8000 Host Attachment and Interoperability 10 7 Working with VERITAS Volume Manager on HP UX When working with HP UX 11i 3 you can choose from two volume managers gt HP Logical Volume Manager LVM gt VERITAS Volume Manager VxVM I O is handled in pass through mode and executed by native multipathing not by DMP According to HP both volume managers can coexist on an HP UX server You can use both simultaneously on separate physical disks but usually you will choose one or the other and use it exclusively For more information see the HP UX System Administrator s Guide Overview HP UX 11i Version 3 at this website http bizsupport1 austin hp com bc docs support SupportManual c02281492 c02281492 pdf For the configuration of the DS8000 logical volumes on HP UX with LVM see 10 6 1 HP UX multipathing solutions on page 187 Example 10 11 shows the initialization of disks for VxVM use and the creation of a disk group with the vxdiskadm utility Example 10 11 Disk initialization and disk group creation with vxdiskadm vxdisk list DEVICE TYPE DISK GROUP STATUS c2t0d0 auto none online invalid c2t1d0 auto none online invalid cl0t0d1 auto none
118. 85760 20062 9811599 1 03461 dev vx dsk 20781_dg 20781_vol 11534336 20319 10794398 1 20781 To expand the volume on the DS8000 use the DS CLI chfbvol command as shown in Example 9 14 Example 9 14 Expanding a volume dscli gt chfbvol cap 18 4704 IBM 2107 7520781 CMUC00332W chfbvol Some host operating systems do not support changing the volume size Are you sure that you want to resize the volume y n y CMUCO00261 chfbvol FB volume 4704 successfully modified The new capacity must be larger than the previous one you cannot shrink the volume To check that the volume was expanded use the 1sfbvol command as shown in Example 9 15 Example 9 15 Isfbvol after expansion dscli gt Isfbvol 4704 Name ID accstate datastate configstate deviceMTM datatype extpool cap 2 30B cap 10 9B cap blocks ITSO_v880_4704 4704 Online Normal Normal 2107 900 FB 512 P53 18 0 37748736 In Example 9 15 the volume 4704 is expanded to 18 GB in capacity To see the changed size of the volume on the Solaris host after the expansion use the luxadm command Example 9 16 Example 9 16 luxadm output after volume expansion bash 3 00 luxadm display dev rdsk c2t50050763030CC1A5d0s2 DEVICE PROPERTIES for disk dev rdsk c2t50050763030CC1A5d0s2 Status Port A 0 K Chapter 9 Solaris considerations 169 Vendor IBM Product ID 2107900 WWN Node 5005076303ffcla5 WWN Port A 50050763030cc1la5 Revision 991 Serial Num 752078147
119. 9 minutes NOMAX Bottom F3 Exit F4 Prompt F5 Refresh F12 Cancel F13 How to use this display F24 More keys Figure 11 41 Moving data from units marked ENDALC You can specify a time limit that the function is to run for each ASP being balanced or the balance can be set to run to completion If you must end the balance function prior to this use the End ASP Balance ENDASPBAL command A message is sent to the system history QHST log when the balancing function is started for each ASP A message is also sent to the QHST log when the balancing function completes or is ended If the balance function is run for a few hours and then stopped it continues from where it left off when the balance function restarts so that the balancing can be run during off hours over several days To remove the old units from the configuration you must use DST and reload the system or partition with IPL With this method you can remove the existing storage units over a period of time This method requires that both the old and new units are attached to the system at the same time Therefore additional IOPs and IOAs might be required if you are migrating from an ESS to a DS8000 It might be possible in your environment to reallocate logical volumes to other IOAs but careful planning and implementation are required Chapter 11 IBM i considerations 245 246 IBM System Storage DS8000 Host Attachment and Interoperability 12 IBM System z consideration
120. 99 2455 South Road Poughkeepsie NY 12601 5400 Stay connected to IBM Redbooks gt Find us on Facebook http www facebook com IBMRedbooks gt Follow us on Twitter http twitter com ibmredbooks gt Look for us on LinkedIn http www 1linkedin com groups home amp gid 2130806 gt Explore new Redbooks publications residencies and workshops with the IBM Redbooks weekly newsletter https www redbooks ibm com Redbooks nsf subscribe 0penForm gt Stay current on recent Redbooks publications with RSS Feeds http www redbooks ibm com rss html Preface xiii xiv IBM System Storage DS8000 Host Attachment and Interoperability General considerations This chapter provides general information for the host attachment of DS8000 storage systems as well as information that pertains to all chapters of this book The following topics are covered DS8000 general topics Client architecture overview SAN considerations Storage tiers v Yv y Copyright IBM Corp 2011 2013 All rights reserved 1 1 1 DS8000 general topics This introductory chapter addresses some of the considerations that apply to all of the different types of hosts that can be attached to a DS8000 storage system It also contains information covering many of the conventions and names used throughout the rest of this book The DS8000 family of storage systems has grown through the years steadily adding new features and industry leading perform
121. Attaching Solaris to DS8000 Solaris uses the LUN polling method to discover DS8000 LUNs For this reason each Solaris host is limited to 256 LUNs per HBA and volume group from the DS8000 IBM offers a mechanism to overcome this limitation This workaround provides the ability to map multiple LUN groups of 256 Using the help mkhostconnect DSCLI command presents the output as shown in Example 9 3 Example 9 3 Extract of the help mkhostconnect DSCLI command output wwname wwpn Required Specifies the worldwide port name WWPN The WWPN is a 16 character hexadecimal ID The names are host attachment specific for example 12341234000A000F Note You should not create more than one hostconnect per WWPN except for SUN hosts Creating more than one hostconnect per WWPN each with a separate volume group is only supported for SUN Setting up this approach requires an approved SCORE it is a special IBM release process to get a formal approval from development labs for not generally supported environments For more information about SCORE see 2 3 Additional supported configurations on page 13 Chapter 9 Solaris considerations 161 You can assign LUNs using any of the supported DS8000 user interfaces including the DS CLI the DS GUI and the DS Open API When using the DS CLI make the host connections using the flags addrdiscovery lunpolling 1bs 512 and profile SUN Solaris Another option is to use the hostty
122. Attachment Guide The IBM System Storage DS8000 Host Systems Attachment Guide SC26 7917 provides information about preparing an Intel IA 32 based machine for DS8000 attachment including the following instructions Working with the unique Linux SCSI subsystem Installing and configuring the Fibre Channel HBA Discovering DS8000 volumes Setting up multipathing Preparing a system that boots from the DS8000 YYYY Y For more information see the BM System Storage DS8000 Host Systems Attachment Guide SC26 7917 at the following website http www ibm com support docview wss uid ssg1S7001161 IBM System Storage DS8000 Host Attachment and Interoperability This publication does not cover other hardware platforms such as the IBM Power Systems or IBM System z Online Storage Reconfiguration Guide The Online Storage Reconfiguration Guide is part of the documentation provided by Red Hat for Red Hat Enterprise Linux 5 and 6 Although written specifically for Red Hat Enterprise Linux 5 most of the general information in this guide is valid for Linux 6 as well This guide covers the following topics for Fibre Channel and iSCSI attached devices Persistent device naming Dynamically adding and removing storage devices Dynamically resizing storage devices Low level configuration and troubleshooting vvvy This guide is available at the following websites gt Red Hat 5 http docs redhat com docs en US Red_Hat_Enterprise_Linux 5 html O
123. B Figure 7 20 Volume properties d Select the LUN notice the new size as shown in Figure 7 21 eater Name Identifier Path ID LUN Capacity Expandable or VMFS Label Name Path 1D LUN Drive Type IBM Fibre Channel Disk naa 60050 wmhba2 C0 TO L1 1 Non SSD mplete Figure 7 21 Select LUN e Review the current disk layout as in Figure 7 22 and ensure that it only contains free space besides the already existing VMFS partition Chapter 7 VMware vSphere considerations 133 Review the current disk layout 6k Layout Device Drive apacit Available mplete IBM Fibre Channel Disk naa Non SSD 40 00 GB 20 00 GB 1 vmfs devices disks naa 600507630affc29f000000000000a000 GPT Free space 1 00 MB y VMFS IBM Fibre Channel Disk naa 20 00 GB Vv Free space 20 00 GB There is only one layout configuration available Use the Next button to proceed with the other wizard pages Free space will be used to expand the VMFS volume Figure 7 22 Disk layout f Select all of the available capacity to be added to the datastore Figure 7 23 Capacity Layout Maximum available space mplete C Custom space setting 20 00 of 20 00 GB available space Figure 7 23 Capacity to add g Review the summary page as in Figure 7 24 and click Finish E Review the proposed disk layout for the new
124. Channel adapters before installation 240 IBM System Storage DS8000 Host Attachment and Interoperability In the IBM i profile the virtual Fibre Channel adapter ID 14 is tagged as a load source Because the installation is done from the images of DVD drives in the VIOS virtual repository the virtual SCSI adapter connecting the repository is tagged as the alternate restart device as shown in Figure 11 35 EROAA https 9 155 57 94 hme wel T2ac7 Logical Partition Profile Properties default_npiv ATS_LPAR19_ITSO_JANA_IP42 p6 570 2 O6C6DE1 ATS_LPAR19_ITSO_JANA_IP42 General Processors Memory I O rual E Controlling Settings Ethernet 1 0 OptiConnect Adapters LHEA Tagged I O devices for this partition profile are detailed below Load source Client Fibre Channel Slot 14 Alternate restart device Client SCSI Slot 2 Console Hardware Management Console Alternate console None Operations Console Figure 11 35 Tagged virtual adapters Chapter 11 IBM i considerations 241 The IPL source must be specified as D under the Settings tab in the Partition Properties window Figure 11 36 x3550 lab 8 Properties Mozilla Firefox IBM Edition BAX EE htys 9 155 57 94 hmc wd T2cc9 i Partition Properties ATS_LPAR19_ITSO_JANA_IP42 General Hardware Virtual Adapters Settings Boot IPL source il Keylock position eae ha Automa
125. DEVICE IS A HYPERPAV ALIAS HyperPAV system requirements HyperPAV has the following z OS requirements gt HyperPAV is supported starting with z OS V1 8 gt For prior levels of the operating system the support is provided as an SPE back to z OS V1 6 as follows IOS support APAR OA13915 DFSMS support including DFSMS SMS the always on management AOM command and the DEVSERV command with APARs 0A13928 0A13929 OA14002 OA14005 OA17605 and OA17746 WLM support APAR OA12699 GRS support APAR OA14556 ASM support APAR OA14248 gt RMF APAR OA12865 In addition DS8000 storage systems require the following licensed functions gt FICON attachment gt PAV gt HyperPAV Chapter 12 IBM System z considerations 267 Important Make sure to have the latest z OS APAR OA39087 Media Manager QSAM BSAM use of ZHPF must remain or be disabled until fixes for OA38939 and 0A39130 and prerequisites OA38966 and OA38925 prior to installing microcode at release 6 2 or higher For more details visit the link http www ibm com support docview wss uid ssg1S1004024 12 5 4 Extended Distance FICON For the DS8000 Extended Distance FICON also known as Simplified FICON Channel Extension is an enhancement to the FICON architecture It can eliminate performance degradation at extended distances having no channel extender installed by implementing a new information unit IU pacing protocol You can use a les
126. DS8000 Host Systems Attachment Guide SC26 7917 at the following website http www 304 ibm com support docview wss uid ssg1S7001161 When attaching an AIX system to the DS8000 series you can distinguish between direct attachment using physical or virtual Fibre Channel adapters in the AIX partition or virtual attachment using virtual SCSI disks which are backed by DS8000 volumes in a VIOS partition Both approaches are described in this chapter The following topics are covered Attaching native Fibre Channel Attaching virtual SCSI Important additional considerations Multipathing with AIX Configuring LVM Using AIX access methods for I O Expanding dynamic volume with AIX SAN boot support vvvvvvvy 61 5 1 Attaching native Fibre Channel To allocate DS8000 disks to AIX the WWPN of each HBA on the host server must be registered in the DS8000 The WWPNs of physical or virtual Fibre Channel adapters can be seen in the output of the 1scfg command as shown in Example 5 1 Example 5 1 List WWPNs of the Fibre Channel adapters in an AIX partition Isdev SA 1 fcs p7 770 02v5 root Isdev SA 1 fcs fcs0 Available 00 00 8Gb PCI Express Dual Port FC Adapter df1000f114108a03 fcs1 Available 00 01 8Gb PCI Express Dual Port FC Adapter df1000f114108a03 Isdev SA 1 fcs while read a b do echo a 1scfg vl a grep Net done fcsO Network Address 10000000C9808618 fcs1 Network Address 1000000
127. Distance FICON 00 0c 268 12 5 5 High Performance FICON for System z with multitrack support ZHPF 269 12 5 6 ZHPF latest enhancements 0 00 eee 270 12 6 z VM considerations ai n inona ea a eet 273 12 6 CONnNeCtVity ceca zewe i tet ogee Aa A T Ea E TE a AET ea E E 273 Contents vii 12 6 2 Supported DASD types and LUNS 000 00 cee 273 12 6 3 PAV and HyperPAV z VM support 0000 0c eee eee 273 12 6 4 Missing interrupt handler 2 0 0 e ee eae 274 12 7 VSE ESA and z VSE considerations 0 0 0 e eee eee 274 12 8 I O Priority Manager for Z OS 0 00 cette 275 12 8 1 Performance groupS 0 eee 275 12 8 2 Easy Tier and I O Priority Manager coexistence 0 000ees 276 12 9 TPC R V5 1 ina z OS environment 0 aaau 278 12 9 1 Tivoli Storage Productivity Center for Replication for System z TPC R 278 12 9 2 References es nefd nee itd oe edad eee bie See ee Es 279 12 10 Full Disk Encryption FDE aaua aaa 280 Chapter 13 IBM SAN Volume Controller considerations 281 13 1 IBM System Storage SAN Volume Controller 00 000 eee eee 282 13 2 SAN Volume Controller multipathing 0 0 00 cece eee 282 13 3 Configuration guidelines for SVC 0 tee 282 13 3 1 Determining the number of controller ports for DS8000 282 13 3 2 Logical volume considerations
128. ER5 Native System i models Yes Yes IBM System p models for IBM i in an LPAR a The support of VIOS NPIV requires IBM i V6 1 1 on either POWER system or BladeCenter b BladeCenters in chassis S support only VIOS VSCSI connection of DS8000 to IBM i Table 11 2 shows the minimal VIOS levels needed for NPIV support VSCSI support and multipath with two VIOS Table 11 2 VIOS levels Methods of attaching POWER system or Minimal VIOS level Minimal VIOS level with VIOS Blade for connection for multipath VIOS NPIV POWER7 V2 1 2 V2 1 2 POWER6 V2 1 2 V2 1 2 POWER Blade V2 2 V2 1 2 P VIOS VSCSI POWER7 V1 5 0 V2 1 2 POWER6 V1 5 0 V2 1 2 POWERS V1 5 0 V2 1 2 POWER Blade V1 5 0 v2 1 2 a BladeCenters in chassis S support only VIOS VSCSI connection of DS8000 to IBM i b Two VIOS in a BladeCenter are supported providing that the BladeCenter is managed by Systems Director Management Console and that there are enough Fibre Channel and Ethernet adapters for each VIOS The following websites provide up to date information about the environments used when connecting DS8800 to IBM i gt System i Storage Solutions http www ibm com systems i hardware storage index html gt Virtualization with IBM i PowerVM and Power Systems http www ibm com systems i os gt IBM Power Systems Hardware Information Center http publib boulder ibm com infocenter powersys v3r1lm5 index jsp to
129. ERS systems that allow an IBM i partition System i models 270 520 525 550 570 595 800 810 820 825 830 840 870 890 and IBM System p models 9117 570 9119 590 and 9119 595 with feature 9411 100 that allow IBM i in a partition Attachment with VIOS NPIV and attachment with VIOS with VSCSI are supported on POWER7 and POWERS6 systems that allow an IBM i partition For more information about POWER systems see the IBM Power Systems Hardware Information Center as listed in 11 1 4 Useful websites on page 204 Attachment of DS8000 with VIOS virtual SCSI or VIOS NPIV is supported for IBM i in POWER6 and POWER based BladeCenters that allow for an IBM i LPAR BladeCenters BladeCenters in chassis S support only VIOS VSCSI connection of DS8000 to IBM i while BladeCenters in chassis H provide both VIOS virtual SCSI and VIOS NPIV connection For more information about supported attachments see the BladeCenter Interoperability Guide as listed in 11 1 4 Useful websites on page 204 For more information about the BladeCenters see the JBM Power Blade servers listed in 11 1 4 Useful websites on page 204 11 1 2 Software This section covers the various software levels IBM i levels The terminology of BM i level is Version Release for example V7R1 or V7 1 stands for Version 7 Release 1 Before IBM i V7 1 the software levels between the releases were denoted by modification levels for example V6 1 1
130. ESS Provider Configuration Tool Commands ibmvssconfig exe lt command gt lt command arguments gt Commands h help showcfg listvols lt all free vss unassigned gt add lt volumeID list gt separated by spaces rem lt volumeID list gt separated by spaces Configuration set targetESS lt 5 digit ESS Id gt set user lt CIMOM user name gt set password lt CIMOM password gt set trace 0 7 set trustpassword lt trustpassword gt set truststore lt truststore location gt set usingSSL lt YES NO gt set vssFreeInitiator lt WWPN gt set vssReservedInitiator lt WWPN gt set FlashCopyVer lt 1 2 gt set cimomPort lt PORTNUM gt set cimomHost lt Hostname gt set namespace lt Namespace gt With the ibmvssconfig exe listvols command you can also verify what volumes are available for VSS in the VSS_FREE pool as shown in Example 3 13 Example 3 13 VSS list volumes at free pool C Program Files IBM ESS Hardware Provider for VSS gt ibmvssconfig exe listvols free Listing Volumes LSS Volume Size Assigned to 10 003AAGXA 5 3687091E9 Bytes5000000000000000 11 103AAGXA 2 14748365E10 Bytes5000000000000000 Also disks that are unassigned in your disk subsystem can be assigned with the add command to the VSS_FREE pool Example 3 14 lists the volumes available for VSS Example 3 14 VSS list volumes available for VSS C Program Files IBM ESS Hardware Provider for VSS gt ibmvssconfig exe listvols vss
131. FS Healthy 465GB 462GB 99 No H Fa Shared Folders amp D58000_1001 G Simple D NTFS Healthy 465GB 462GB 99 No E Local Users and Groups Performance Logs and Alerts z Device Manager Storage Ep Removable Storage Disk Defragmenter Disk Management l T Services and Applications ae amp SDisk 0 SaaS Basic c 33 89 GB 33 89 GB NTFS Online Healthy System SDisk 1 Sa een Basic 33 90 GB 33 90 GB Online Unallocated lt BDisk 2 Basic ds8000_1000 E 4 65 GB 4 65 GB NTFS Online Healthy Disk 3 Dynamic DS8000_1001 G 4 65 GB 4 65 GB NTFS Online Healthy Figure 3 3 Disk Management view with two devices Tip Newly assigned disks will be discovered if not go to the Disk Manager window and rescan the disks or go to the Device Manager window and scan for hardware changes 3 2 1 SDD datapath query When using the datapath device query command option 1 is added to mark the non preferred paths with an asterisk For example with datapath query device 1 all paths of the DS8000 are used because the DS8000 does not implement the concept of preferred path as the SAN volume controller SVC does Example 3 1 shows the output of the datapath query device 1 command Example 3 1 Output of datapath query device I command Microsoft Windows Version 5 2 3790 C Copyright 1985 2003 Microsoft Corp C Program Files IBM Subsystem Device Driver gt da
132. I FCP 0 These I O ports must match the ports listed by the Ishostconnect login command as shown in Example 4 3 on page 51 For a proper assignment of the DS8000 volumes as VSCSI backing devices on the VIOS to the client LPARs running on AIX IBM i or Linux on POWER correctly identify the association of DS8000 volumes with VIOS hard disks Each VIOS hdisk can be associated with the DS8000 volume ID by using the command shown in Example 4 8 Example 4 8 Associate VIOS hdisks to DS8000 volumes Isdev type disk grep 2107 while read a b do echo a 1sdev dev a vpd grep Serial done hdisk0 Serial Number 75TV1819000 hdiskl Serial Number 75TV1819004 hdisk2 Serial Number 75TV1819005 In redundant VIOS environments such as dual VIOS environments it is important to uniquely identify the hdisks by their DS8000 serial number to assign the correct volumes on both VIOS to the client partitions for proper multipathing Another alternative to uniquely identify the hard disks on the VIOS is to use the chkdev command which displays the UDID of each hdisk as shown in Example 4 9 Example 4 9 Identifying DS8000 LUNs on VIOS using UDID chkdev grep E NAME ID NAME hdiskO IDENTIFIER 200B75TV1819000072107900031BMfcp NAME hdiskl IDENTIFIER 200B75TV1819004072107900031BMfcp NAME hdisk2 IDENTIFIER 200B75TV1819005072107900031BMfcp In a dual VIOS environment with two virtual I O server
133. I Os Reducing elapsed time by reducing the wait time in channel path recovery Read availability mask support The dynamic channel path identifier management DCM provides you with the ability to define a pool of channels that are managed by the system The channels are added and deleted from control units based on workload importance and availability needs DCM attempts to avoid single points of failure when adding or deleting a managed channel by not selecting an interface on the control unit on the same I O card Initial program load enhancements During the Initial Program Load IPL sequence the channel subsystem selects a channel path to read from the system residence volume SYSRES device Certain types of I O errors on a channel path cause the IPL to fail even though there are alternate channel paths that might work For example consider a situation where you have a bad switch link on the first path but good links on the other paths In this case you cannot IPL because the same faulty path is always chosen The channel subsystem and z OS are enhanced to retry I O over an alternate channel path This circumvents IPL failures that are caused by the selection of the same faulty path to read from the SYSRES device 12 3 2 DS8000 device definition 250 To exploit the increase in the number of LSSs that can be added in the DS8000 255 LSSs the unit must be defined as 2107 in the Hardware Configuration Definition HCD or input outpu
134. IBM Fibre Channel Disk naa 20 00 GB Disabled Datastore Name D50000 Da J Formatted Capacity 19 75 GB Paths 7 Total 8 Formatii koa o File System VMFS 5 54 Disabled 0 ics Figure 7 16 Datastore details Fixed LUN ID for ESXi volumes For certain scenarios it is required that the same volume is represented with the same LUN or SCSI ID on all hosts The LUN ID represents the position on the SCSI bus By default the DS8000 does not specify this ID and the ESXi host attaches the volumes in the order they appear starting with 0 It might lead to two ESXi hosts seeing the same volume on different positions in the SCSI bus In Example 7 1 volume A005 is assigned to two hosts While host 1 sees the volume in position 00 host 2 sees it in position 01 Example 7 1 Different LUN ID for two hosts dscli gt showvolgrp lunmap v64 Date Time 1 Juni 2012 12 41 00 CEST IBM DSCLI Version 7 6 20 221 DS 75 TV181 Name ITSO ESX 1 ID V64 Type SCSI Map 256 Vols A005 Mapping vol lun A005 00 dscli gt showvolgrp lunmap v65 Date Time 1 Juni 2012 12 41 02 CEST IBM DSCLI Version 7 6 20 221 DS 75T V181 Name ITSO ESX 2 ID V65 Type SCSI Map 256 Vols A000 A005 130 IBM System Storage DS8000 Host Attachment and Interoperability A000 00 A005 01 To avoid a situation where two hosts see the same LUN with different SCSI IDs it is possible to specify the SCSI ID when assigning the vo
135. IBM System Storage DS8000 Host Attachment and Interoperability Learn how to attach DS8000 to open systems IBM System z and IBM i See how to gain maximum availability with multipathing Discover best practices and considerations for SAN boot Axel Westphal Bertrand Dufrasne Juan Brandenburg Jana Jamsek Kai Jehnen Steven Joseph Massimo Olivieri Ulrich Rendels Mario Rodriguez Redbooks ibm com redbooks International Technical Support Organization IBM System Storage DS8000 Host Attachment and Interoperability February 2013 SG24 8887 01 Note Before using this information and the product it supports read the information in Notices on page Ix Second Edition February 2013 This edition applies to the IBM System Storage DS8000 with licensed machine code LMC level 6 6 xxx xx bundle version 86 x xxx xx and LMC level 7 6 xxx xx Copyright International Business Machines Corporation 2011 2013 All rights reserved Note to U S Government Users Restricted Rights Use duplication or disclosure restricted by GSA ADP Schedule Contract with IBM Corp Contents Notices opere feiss AE ee Lee ele as Sieben ee ix PACS IMARKS 5 atu Seek AN a ath A IR ns ARN IN Selects Me aN A aea aa tics tlk Me ERG ans at x Preface a a neti aie ee edt eee Ga EA Gn ae aed kee a xi The team who wrote this book 2 0 0 ete xi Now you can become a published author too 2 0000 eee xiii Comme
136. KB of disk space in volume group version 1 0 On volume group version 2 x the physical volume size will be recalculated to use full size upto a maximum of 17179869184kB Physical volume dev rdisk disk1078 has been successfully created mknod dev vgMZOutCTest010 group c 128 0x0a0000 vgcreate V 2 1 s 256 S 32t dev vgMZOutCTest010 dev disk disk1078 Volume group dev vgMZOutCTest010 has been successfully created Volume Group configuration for dev vgMZOutCTest010 has been saved in etc lvmconf vgMZOutCTest010 conf Set the following parameters a s 256 lowercase small s sets the Physical Extents size to 256 the default is 4 b S 32t uppercase large S sets the Maximum Volume Group size to 32t Tip Upon entering the vgcreate command if you specify a major number such as 64 you see the following message vgcreate Error The major number for the group file dev vgMZOutCTest010 group corresponds to volume group version 1 0 Create the logical volumes and the corresponding file systems Requirement A license is required to work with large lvols If you do not have a license you see the following message UX vxfs mkfs ERROR V 3 26141 mkfs You don t have a license to create a file system of size gt 2147483648 sectors 2048 GB Chapter 10 HP UX considerations 199 200 IBM System Storage DS8000 Host Attachment and Interoperability 11 IBM i considerations This chapter provides the specifics
137. Linux on System z provides the kernel module zfcp to operate the FICON adapter in FCP mode A channel card can only run either in FCP or FICON mode Also in FCP mode it must be dedicated to a single LPAR and cannot be shared To maintain redundancy use more than one FCP adapter to connect to the DS8000 volumes Linux will see a separate disk device for each path and requires DM Multipath to manage them Linux on System z running in a virtual machine under z VM Running several of virtual Linux instances in a z VM environment is much more common The IBM z VM operating system provides granular and flexible assignment of resources to the virtual machines VMs and also enables sharing of resources between VMs 88 IBM System Storage DS8000 Host Attachment and Interoperability The z VM offers several ways to connect storage to its VMs such as the following methods gt Channel attached DASD device The z VM system can assign DASD devices directly to a virtual machine If a VM runs Linux it can access the device with the DASD driver Multipathing is handled in the I O subsystem therefore DM Multipath is not required PAV is supported in the same method as for Linux running directly in an LPAR gt Z VM minidisk A minidisk is a virtual DASD device that z VM models from a storage pool it has assigned for its use It can be attached to a VM Within the VM it is treated as though it were a real device Linux uses the DASD driver for acc
138. Load source U789D 001 DQD904G P1 C2 Fibre Channel Serial Bus H Hardware Management Console Alternate console None Operations Console None Figure 11 29 HMC Select the adapter that connects external Load Source When IBM i resides on a Blade server and the DS8000 is connected with VIOS NPIV or Virtual SCSI the load source adapter defaults to the single client adapter that is initially in the IBM i partition Chapter 11 IBM i considerations 237 Figure 11 30 shows a partition to be created with Integrated Virtualization Manager IVM on a Blade server After the partition is created you can add other virtual adapters and change the tagging of the load source on the partition properties page in IVM Create Partition Summary Step 8 of 8 Ff Summary This is a summary of your partition settings Select Finish to create the partition To make changes to the settings select Back You can modify the partition by using the partition properties task after you complete this wizard System name Server SN Partition ID 3 Partition name iSOSTest Environment iS OS Memory 1 GB 1024 MB Processors 1 virtual Virtual Ethernets i 1 Host Ethernet adapter ports None Storage capacity 120 GB 122880 MB hdisk26 Storage devices hdisk27 hdisk28 _ hdisk29 Optical devices None Virtual optical devices 1 Physical adapters None Alternate res
139. N evaluations and upgrade planning 1 4 Storage tiers This publication has several references to drive tiers storage tiers and the IBM Easy Tier application This section discusses what is meant by the term tier relative to the DS8800 When a workload for storage is laid out or provisioned there are some types of data that are known ahead of time to be more resource hungry That is the data from some workloads will be accessed more frequently or might change more frequently than other workloads Some applications require data to be stored and then only occasionally read other applications require thousands of reads and writes daily Chapter 1 General considerations 3 4 Because of the differing needs of storage data the DS8800 supports installation of three physically different classes of storage all of which use the term Disk Drive Module DDM The real world capabilities of these storage types are different enough that they are considered to be in different classes of performance also known as tiers Here are the types gt Solid State SSD NAND based flash memory is used for storage no moving parts 2 5 form factor 300 GB to 400 GB considered as the highest tier gt Enterprise Traditional electromechanical drive high 10 000 and 15 000 RPM 2 5 form factor 146 GB to 900 GB considered as a middle tier gt Nearline Traditional electromechanical drive 7 200 RPM 3 5 form factor 3 TB considered as the
140. N Port_id is None N_Port Node World Wide Name 0x5001438001321d79 N_Port Port World Wide Name 0x5001438001321d78 Switch Port World Wide Name 0x201d00051e349e2d Switch Node World Wide Name 0x100000051e349e2d Driver state ONLINE Hardware Path is 0 3 1 0 Maximum Frame Size 2048 Driver Firmware Dump Available NO Driver Firmware Dump Timestamp N A Driver Version fcd B 11 31 0709 Jun 11 2007 184 IBM System Storage DS8000 Host Attachment and Interoperability The HP UX device file for the HBA dev fcdX is a required parameter Example 10 3 on page 184 shows that the HBAs are dual port devices which is why you see two WWPNs for each HBA 10 5 Configuring the HP UX host for the DS8000 The first step in configuring a host connection on the DS8000 is to define a host connection using the graphical interface DS GUI or the command line interface DS CLI The hosttype parameter is required for this definition Predefined hosttype HP automatically configures the DS8000 to present the DS8000 volumes in an HP UX preferred method Keep in mind that when you create or assign LUNs and volumes only LUN and volume IDs less than x 3FFF are recognized in HP UX 11i v3 as mentioned in Notes on enumeration on page 183 10 5 1 Device special files In HP UX 11i device special files DSF appear in the dev directory like regular files but they are special They are associated with either real physical devices or pseudo de
141. N WWPN ESSIOport LoginType Name ID 281E000C2900000C 281E000C2900000D 10001 SCSI ITSO ESXi NPIV 0075 281E000C2900000C 281E000C2900000D 10005 SCSI ITSO ESXi NPIV 0075 Tip If the VM is not able to access the RDM using its virtual WWN it will access the LUN using the ESXi hosts WWN as if NPIV was not configured 7 3 Multipathing The following sections provide an insight into the way VMware handles multiple paths to a storage device 7 3 1 Pluggable Storage Architecture 142 For multipathing ESXi uses a concept called Pluggable Storage Architecture PSA As shown in Figure 7 40 it consists of three levels of plug ins gt Native Multipathing NMP This plug in chooses where to send I O requests and handles the different storage devices connected to the ESXi host gt Storage Array Type Plug in SATP This plug in is responsible for array specific operations such as path monitoring and activating SATP is chosen by NMP based on the type of storage device detected The user is able to change the SATP by esxcli commands gt Path Selection Plug in PSP This plug in is responsible for choosing the physical path for an I O request NMP assigns PSP to SATP but the user can change the PSP concurrently by the GUI Tip The PSP can be chosen individually for each LUN IBM System Storage DS8000 Host Attachment and Interoperability Pluggable Storage Architecture PSA VMware NMP VMware SATP
142. O0DE908 136 62 GB sj Device Mapper dev mapper SServeRA _Drive_1_2DO0DE908 partl 39 99 GB Q NFS dev mapper SServeRA_Drive_1_2DODE908_part2 39 99 GB dev mapper SServeRA_Drive_1_2DODE908_part3 8 00 GB R Unused Devices dev sda e z 136 62 GB settings idevisdb 50 00 GB R Figure 6 6 Selecting multipath device for installation 5 Select the multipath device DS8000 volume to install to and click Accept 6 On the Partitioner panel create and configure the required partitions for your system in the method you use on a local disk 6 5 8 Installing with YAST You can also use the automatic partitioning capabilities of YAST after the multipath devices were detected Follow these steps 1 Click the Back button until you see the Preparing Hard Disk Step 1 panel Figure 6 4 on page 120 2 When you see the multipath devices instead of the disks select the multipath device you want to install on and click Next 3 Select the partitioning scheme Important All supported platforms can boot Linux from multipath devices In certain cases however the tools that install the boot loader can only write to simple disk devices You will need to install the boot loader with multipathing deactivated SLES10 and SLES11 provide this capability with the parameter mul tipath off to the boot command in the boot loader The boot loader for IBM Power Systems and System z must be installed again when the kernel or InitRAMFS i
143. OS V1R11 V1R12 V1R13 and above ZHPF Pre fetch Optimizer LPO In addition to ZHPF list prefetch R6 2 introduced a new caching algorithm called List Pre fetch Optimizer LPO List Prefetch Optimizer requires ZHPF Unlike ZHPF list prefetch LPO is internal to the DS8000 code Whereas the objective of ZHPF list prefetch is to reduce the I O connect time the objective of LPO is to reduce disconnect time Release level 6 3 went further by fine tuning LPO for some types of list prefetch I O on ZHPF Prefix and locate commands can provide a list of records CCHHR to be read by the next locate operation in the command chain gt ZHPF allows the control unit to anticipate the next set of records while the host is reading the current set of records gt ZHPF list of up to 21 records per command gt Supported on z OS V1R11 V1R12 V1R13 and above gt For more information see the Redpaper publication DB2 for z OS and List Prefetch Optimizer REDP 4682 Chapter 12 IBM System z considerations 271 272 CKD Event Aggregation Summary Starting from Release 6 2 LMC level 7 6 20 xx DS8000 has implemented the CKD Event Aggregation Summary which works as follows gt Reduces host interrupts by up to 1 256 Now the z OS host receives a single notification of each event that affects all volumes in an LSS gt Summary event notification for PPRC suspension events Only if the host supports PPRC suspension summary events each LSS
144. PCM 3 0 0 0 or later releases the command to start the server is startpcmsrv gt For SDDPCM 2 6 0 1 or prior releases the command to start the server is startsrv s pcmsrv e XPG_SUS_ENV ON It is possible as well to stop the SDDPCM server manually as follows gt For SDDPCM 3 0 0 0 or later releases the command to stop the server is stoppcmsrv gt For SDDPCM 2 6 0 1 or prior releases the command to stop the server is stopsrc s pcmsrv For more details about the SDDPCM multipath driver see the Multipath Subsystem Device Driver User s Guide GC52 1309 03 5 5 Configuring LVM This section provides information and considerations when configuring LVM on AIX servers In AIX all storage is managed by the AIX LVM It virtualizes physical disks to dynamically create delete resize and move logical volumes for application use With AIX DS8000 logical volumes appear as physical SCSI disks 5 5 1 LVM striping Striping is a technique for spreading the data in a logical volume across several physical disks in such a way that all disks are used in parallel to access data on one logical volume The primary objective of striping is to increase the performance of a logical volume beyond that of a single physical disk LVM striping can be used to distribute data across more than one array or rank in case of single rank extent pools 76 IBM System Storage DS8000 Host Attachment and Interoperability 5 5 2 Inter physical volume alloca
145. PCM file set to be installed is dependent on the VIOS release VIOS 1 x based on AIX 5 3 or VIOS 2 x based on AIX 6 1 For more information about SDD SDDPCM and file set downloads see the support matrix at this website http www 01 ibm com support docview wss rs 540 amp context ST52G7 amp dc DA4008uid ssg1S 7001350 Also see the readme file of the particular SDD or SDDPCM version for latest updates and prerequisites New SDD or SDDPCM releases typically require a specific minimum version of the host attachment file set For supported SDD and SDDPCM releases with IBM System Storage DS8000 series environments and VIOS releases see the IBM SSIC at this website http www 03 ibm com systems support storage ssic interoperability wss The SDD or SDDPCM installation and management on the VIOS is similar to the description given in Chapter 5 AIX considerations on page 61 You can install file sets on the VIOS by using one of the following commands gt The cfgassist command within the restricted padmin environment gt The oem_setup _env command to gain root privileges and access the OEM install and setup environment as on an AIX system To access the SDD datapath command or the SDDPCM pcmpath command access the OEM install and setup environment using oem_setup_env first These commands provide detailed and compact information about the devices and their paths including device statistics the DS8000 serial number and the DS8000 v
146. Rg Basic Operations zz Work Management E E E pe Configuration and Service i 9 155 70 15 y5r4m0 AS4Q0PFE r nays 9 155 70 32 v rimO iSPFE a sian t il 9 155 70 40 v7rim0 Manage this system din Mcei5l1 mai v5r4m0 Manage this system ll All Hardware Communications ha System Adapters LAN Resources a Workstation Resources Processor Information i Cryptography Resources Optical Units E Disk Units ta Devices Software H Fixes Inventory FA Collection Services fi Logical Partitions a hy Natuarl a Add a connection Add a connection il Change all of your system passwords Install plug ins Figure 11 15 System i Navigator Disk Units Chapter 11 IBM i considerations 229 4 Inthe Service Tools Sign on window Figure 11 16 enter your Service tools ID and password Then click OK Service Tools Sign on _p6 570 lab 2c mainz de ibm Service tools user QSECOFR Password a OOO Figure 11 16 SST Sign on 5 Right click Disk Pools and select New Disk Pool 6 In the New Disk Pool wizard on the window that opens click Next 7 In the New Disk Pool window Figure 11 17 complete these steps a For Type of disk pool select Primary b Enter a name for the new disk pool c For Database leave the default setting Generated by the system d Select the method that matches the type of logical volume you are adding If you do not select one o
147. SSD with 300GB or 400 GB capacities From a performance point of view they can be the best option for heavy I O applications Tip SSD can only be used with RAID 5 and there are other restrictions regarding the use of SSDs For more information about implementing SSD with IBM i see DS8000 Introducing Solid State Drives REDP 4522 Consider using RAID 10 for an IBM i workload because it helps provide significant high availability and performance improvements compared to RAID 5 or RAID 6 11 3 3 Cache The DS8000 cache size is significant for an IBM i workload Use the Disk Magic tool to model the disk response times and DS8000 utilizations for a particular IBM i workload with certain cache sizes and determine the best size of DS8000 cache As a rough guideline for the cache size based on the DS8000 capacity see Table 6 88 Cache guidelines in the Redbooks publication DS8800 Performance Monitoring and Tuning SG24 8013 00 208 IBM System Storage DS8000 Host Attachment and Interoperability 11 3 4 Number of ranks When considering the number of ranks take into account the maximum disk operations per second including RAID penalty and per rank These values are measured at 100 DDM utilization with no cache benefit and with an average I O of 4 KB Larger transfer sizes reduce the number of operations per second The maximum disk operations per second depends on the type of disks being used in the DS8000 series however they
148. TOC used for LVs The size of an EAV VTOC index is increased four times and now has 8192 blocks instead of 2048 blocks Because no space is left inside the Format 1 DSCB new DSCB formats Format 8 and 9 were created to protect existing programs from encountering unexpected track addresses These DSCBs are called extended attribute DSCBs Format 8 and 9 DSCBs are new for EAV The existing Format 4 DSCB was changed also to point to the new Format 8 DSCB Tip VTOC size should be increased not only for INDEX size It is possible only if there is free space behind the VTOC location The VTOC and VTOC index must not have extends Volume size When formatting a volume with ICKDSF where the VTOC index size was omitted the volume takes the default size of 15 tracks 12 4 1 Identifying an EAV When a volume has more than 65 520 cylinders the modified Format 4 DSCB is updated to x FFFE which has a size of 65 534 cylinders It will identify the volume as being an EAV An easy way to identify an EAV is to list the VTOC Summary in Time Sharing Option TSO of the Interactive System Productivity Facility ISPF version 3 4 Example 12 2 shows the VTOC summary of a 3390 Model 3 volume Example 12 2 VTOC information for 3390 3 VTOC Summary Information Volume RE77BE Command gt Unit 3390 Volume Data VTOC Data Free Space Tracks Cyls Used 0 Used 1 Largest 16 53
149. V in BladeCenter requirements gt 8252 QLogic Ethernet and 4 GB Fibre Channel Expansion Card CFFh gt 8251 Emulex 4 GB Fibre Channel Expansion Card CFFv gt 8248 QLogic 4 GB Fibre Channel Expansion Card CFFv 206 IBM System Storage DS8000 Host Attachment and Interoperability CFFv CFFh and ClOv CFF v stands for combination form factor vertical CFFh stands for combination form factor horizontal and ClOv stands for combination I O form factor vertical Up to 16 LUNs can be assigned to a virtual SCSI adapter in IBM i when connecting DS8000 series to an IBM i client SCSI command tag queuing is supported by IBM i with a VIOS VSCSI connection and queue depth 32 input and output operations 11 2 4 Overview of the number of LUNs per adapter and the queue depth Table 11 3 provides an overview of the number of LUNs and supported queue depth on adapters at different connection methods Table 11 3 Maximal number of LUNs per port and Queue depth Connection Max number of Max number of Max number of Queue depth to method LUNs per port LUNs per virtual LUNs per virtual a LUN or max number FC adapter SCSI adapter per two or more ports in multipath Native with 32 0 IOP based adapters Native with 64 63 IOP less adapters VIOS_NPIV 64 6 VIOS Virtual 16 32 SCSI a IBM i level v6 1 or later is needed to support SCSI command tag queuing 11 3 Sizing and implementation guidelines When sizing DS8000
150. V181 the DS8000 volume ID 9004 and the device type 2107 900 Without viewing the virtual device mapping on the VIOS you can see which DS8000 LUN is associated with each virtual SCSI disk device on an AIX client partition Chapter 5 AIX considerations 65 The default AIX MPIO path control module PCM is used on the AIX client partition even though SDD or SDDPCM is used as multipath driver on the VIOS partition with a default multipath algorithm of load balancing for the DS8000 LUN The only available multipath algorithm here is failover as shown in Example 5 7 So for any given Virtual SCSI Disk device only one path to one VIOS is used for I O at a given time even when both paths are enabled AIX MPIO will switch to the secondary path only in the case of a failure of the primary path for example during a shutdown of the connected VIOS Example 5 7 Multipath algorithms for Virtual SCSI Disk devices with default MPIO PCM in AIX Isattr R1 hdiskO a algorithm fail_over In a redundant dual VIOS environment the virtual SCSI disk device is provided by two virtual I O servers to the client partition The I O can be balanced manually between both VIOS partitions by setting the AIX path priority accordingly on the AIX client partition for each virtual SCSI device using the chpath command as shown in Example 5 8 The default path priority is 1 which is the highest priority By default every path has priority 1 and the first one will be used for
151. WNs CPU MMU Virtualization Automatic Number of WWNNs hl Swapfile Location Use default settings Number of WWPNs 1 X WWN Assignments No WWNs currently assigned a Figure 7 38 NPIV options tab 3 Click OK to close the Virtual Machine Properties window The WWNs are now created 4 Reopen the Virtual Machine Properties select the Options tab click Fibre Channel NPIV and note the WWNs for the VM as shown in Figure 7 39 WWN Assignments Node WWN a 28 1e 00 0c 29 00 00 0c Port WWN 28 1e 00 0c 29 00 00 0d Figure 7 39 WWN assignments of a VM Chapter 7 VMware vSphere considerations 141 5 Create zoning for the virtual machines WWPN s 6 Create host connection and volume group for the virtual machines WWPN s using hosttype VMware Important gt The RDM volume needs to be mapped to all ESXi WWPNs that can access the DS8000 and in addition to the virtual machines WWPN s gt If separate volume groups are used for the ESXi host and the virtual machine the RDM volume must have the same SCSI ID in both volume groups gt See Fixed LUN ID for ESXi volumes on page 130 for details 7 When the VM is powered on it will log in to the DS8000 with its virtual WWPN To check use the DS CLI command Ishostconnect login See Example 7 4 Example 7 4 VM logged in with NPIV dscli gt Ishostconnect login Date Time 1 Juni 2012 11 58 03 CEST IBM DSCLI Version 7 6 20 221 DS 75 TV181 WWN
152. _viol_0 0019 10000000C9998F6E pSeries IBM pSeries AIX 60 V15 all p770_viol_1 001A 10000000C9998F6F pSeries IBM pSeries AIX 60 V15 all dscli gt Isfbvol volgrp v15 Name ID accstate datastate configstate deviceMIM datatype extpool cap 2 30B cap 10 9B cap blocks p770_viol 9000 Online Normal Normal 2107 900 FB 512 P8 50 0 104857600 p770_viol 9001 Online Normal Normal 2107 900 FB 512 P8 50 0 104857600 p770_viol 9002 Online Normal Normal 2107 900 FB 512 P8 50 0 104857600 dscli gt Ishostconnect login WWNN WWPN ESSIOport LoginType Name ID Chapter 4 Virtual I O Server considerations 51 20000000C9998F6E 10000000C9998F6E 10005 SCSI p770_viol_0 0019 20000000C9998F6E 10000000C9998F6E 10005 SCSI p770_viol_0 0019 20000000C9998F6F 10000000C9998F6F 10135 SCSI p770_viol_1 001A 20000000C9998F6F 10000000C9998F6F 10135 SCSI p770_viol_1 001A 52 On the VIOS partition either the IBM SDD or the Subsystem Device Driver Path Control Module SDDPCW file sets must be installed for the DS8000 volumes as follows gt IBM SDD devices fcp disk ibm mpio rte devices sdd 61 rte 61 stands for AIX6 1 or VIOS 2 x gt IBM SDDPCM devices fcp disk ibm mpio rte devices sddpcm 61 rte 61 stands for AIX6 1 or VIOS 2 x The SDD or SDDPCM file sets require separate DS8000 host attachment file sets ibm fcp disk ibm as prerequisite You can have only SDD or SDDPCM file sets installed on a given VIOS not both Also note that the SDD or SDD
153. a Figure 11 9 Work with Disk Configuration Select one of the following Display disk configuration Add units to ASPs Work with ASP threshold Add units to ASPs and balance data Enable remote load source mirroring Disable remote load source mirroring Start compression on non configured units Work with device parity protection Start hot spare 10 Stop hot spare 11 Work with encryption 12 Work with removing units from configuration ONO ON BPWDY PF wo Selection 4 F3 Exit F12 Cancel Figure 11 9 Working with Disk Configuration menu IBM System Storage DS8000 Host Attachment and Interoperability 5 In the Add Units to ASPs panel Figure 11 10 select option 1 or 2 to create a new ASP or select option 3 to add LUNs to an existing ASP In this example use option 3 to add the LUNs to the existing ASP1 Add Units to ASPs Select one of the following 1 Create unencrypted ASPs 2 Create encrypted ASPs 3 Add units to existing ASPs Selection 3 F3 Exit F12 Cancel Figure 11 10 Creating a new ASP or using existing ASP 6 In the Specify ASPs to Add Units panel Figure 11 11 specify the ASP number to the left of the desired units and press Enter In this example ASP1 was specified for the system ASP Specify ASPs to Add Unit Specify the existing ASP to add each unit to Specify Serial Resource ASP Number Type Model Capacity Name 1 50 210629F 2107 A04 705
154. a part of the operation The selected disks may also be added to a disk group as spares Or they may be added as nohotuses to be excluded from hot relocation use The selected disks may also be initialized without adding them to a disk group leaving the disks available for use as replacement disks More than one disk or pattern may be entered at the prompt Here are some disk selection examples all all disks c3 c4t2 all disks on both controller 3 and controller 4 target 2 c3t4d2 a single disk in the c t d naming scheme xyz_0 a single disk in the enclosure based naming scheme XYZ_ all disks on the enclosure whose name is xyz Select disk devices to add lt pattern list gt all list q cl0t6d0 cl0t6d1 Here are the disks selected Output format Device Name cl0t6d0 c10t6d1 Continue operation y n q default y y You can choose to add these disks to an existing disk group a new disk group or you can leave these disks available for use by future add or replacement operations To create a new disk group select a disk group name that does not yet exist To leave the disks available for future use specify a disk group name of none Which disk group lt group gt none list q default none dg01 There is no active disk group named dg01 Create a new group named dg01 y n q default y Create the disk group as a CDS disk group y n q default y n Use default disk names for these disks y n q defau
155. able 11 4 Maximal host IOPS for HDD ranks RAID rank type Host IOPS at 70 reads Host IOPS at 50 read DDM type IOP less IOP based IOP less IOP based Fibre Channel DDM RAID 5 15 K RPM 827 551 643 428 RAID 10 15 K RPM 1102 706 982 655 RAID 6 15 K RPM 636 424 463 308 SAS DDM RAID 5 15 K RPM 940 627 731 487 RAID 10 15 K RPM 1253 835 1116 744 RAID 6 15 K RPM 723 482 526 351 Table 11 4 shows RAID 10 can support higher host I O rates than RAID 5 However you must consider this against the reduced effective capacity of a RAID 10 rank as compared to RAID 5 RAID 6 is also performs slightly slower than RAID 5 Chapter 11 IBM i considerations 209 Table 11 5 shows the maximal number of host I Os per rank of SSDs for 60 disk utilization The calculation is done for the average of maximum IOPS to an 6 P and 7 P rank taking into account 20 read cache hit and 30 write cache efficiency Most of the implementations will be done with IOP less adapters so this calculation applies to 60 utilization of SSD valid for IOP less connection Table 11 5 Maximal host IOPS for SSD ranks RAID 5 rank type Host IOPS at 70 reads Host IOPS at 50 read 6 P LFF SSD 10 800 8 400 7 P LFF SSD 12 343 9 900 6 P SFF SSD 18 429 14 333 7 P SFF SSD 21 061 16 381 11 3 5 Sizing for SSD and hot data management with IBM i 210 The following sections provide sizing considerations Using t
156. ache to the application s memory For applications that benefit from the operating system cache do not use direct I O because all I O operations are synchronous Direct I O also bypasses the JFS2 read ahead Read ahead can provide a significant performance increase for consecutively accessed files 78 IBM System Storage DS8000 Host Attachment and Interoperability 5 7 Expanding dynamic volume with AIX Starting with IBM DS8000 licensed machine code 5 3 xx xx it is possible to expand a logical volume in size without taking the volume offline Additional actions are required on the attached host to make use of the extra space This section describes the required AIX logical volume manager LVM tasks After the DS8000 logical volume is expanded use the chvg command with the g option to examine all the disks in the volume group to see if they have grown in size If the disks have grown in size then the chvg command attempts to add additional physical partitions to the AIX physical volume that corresponds to the expanded DS8000 logical volume Example 5 21 shows an AIX file system that was created on a single DS8000 logical volume The DSCLI is used to display the characteristics of the DS8000 logical volumes AIX LVM commands show the definitions of volume group logical volume and file systems The available space for the file system is almost gone Example 5 21 DS8000 logical volume and AIX file system before dynamic volume expansion d
157. acity or a different performance characteristic is required for the new virtual disk Capacity Disk Size id j v Disk Provisioning Thick Provision Lazy Zeroed Thick Provision Eager Zeroed Thin Provision Location Store with the virtual machine Specify a datastore or datastore duster Figure 7 29 virtual disk parameters 6 In the advanced options page the default values fit in most usage scenarios Figure 7 30 shows an example with default values Virtual Device Node This node specifies where in the virtual SCSI bus of the VM the new virtual disk should be attached Mode This mode specifies wether the virtual disk is included in snapshots If Independent is selected the virtual disk will not be included in any snapshots For independent disks there are two options e Persistent All data is written directly to the disk and saved across power cycles e Nonpersistent All data written to this disk is temporarily saved and lost after a power cycle Chapter 7 VMware vSphere considerations 137 Specify the advanced options for this virtual disk These options do not normally need to be changed Virtual Device Node o scst 0 1 v C IDE 0 0 Mode Independent Independent disks are not affected by snapshots c Changes are immediately and permanently written to the disk c Changes to this disk are discarded when you power off or revert to
158. adapters consider the capacities per adapter that are shown in Table 11 6 on page 214 and calculate the number of LUNs from them Table 11 9 gives you an overview of the preferred number of LUNs per port for different adapters and different sizes of LUNs The calculation for the table assumes workload Access Density 1 5 Table 11 9 Preferred number of LUNs per port Preferred number of LUNs per port IOP 2844 IOA 5760 4 Gb adapters 5749 5774 8 Gb adapter 5735 35 GB LUNs 24 64 64 70 GB LUNs Size not advisable 39 46 141 GB LUNs Size not advisable 20 23 For two ports in multipath use the 2 times preferred number If it is bigger than the maximal number of LUNs per port use the maximum number per port When connecting DS8000 series to IBM i with VIOS NPIV you can assign 64 LUNs per port in virtual Fibre Channel adapters or 64 LUNs per multiple ports in multipath without a performance impact When connecting DS8000 series to IBM i with VIOS VSCSI you can define up to 16 LUNs per virtual port or up to 16 LUNs per multiple ports in multipath without a performance impact Tip IBM i multipath with VIOS NPIV or VIOS VSCSI is a multipath with two or more VIOS with each path using a separate VIOS 11 4 2 Sharing or dedicating ranks for an IBM i workload Consider using separate extent pools in DS8000 series for IBM i workload and other workloads This method isolates the I O for each server and he
159. age 244 or 11 9 2 IBM i data migration on page 244 11 9 1 Metro Mirror and Global Copy Depending on the existing configuration it might be possible to use Metro Mirror or Global Copy to migrate from an IBM TotalStorage Enterprise Storage Server ESS or any other DS8000 family model to a newer DS8000 model It might also be possible to use any combination of external storage units that support Metro Mirror and Global Copy For more information about Metro Mirror and Global Copy see BM System Storage DS8000 Copy Services in Open Environments SG24 6788 You can use DS8000 copy services for migration when the IBM i has the load source unit and all the other disks on a DS8000 and you plan to migrate the IBM i workload to another DS8000 on 1000 km remote location with the two storage systems connected through dark fiber To do this you must establish a Global Copy relationship with IBM i LUNs at the remote DS8000 When ready to migrate to the remote site you must also do a complete shutdown of the IBM i to ensure all data is written to disk and wait until Global Copy transfers all the updates to the remote DS8000 with no out of synch tracks Additionally unassign the DS8000 LUNs on local site and assign the DS8000 LUNs to the remote IBM i LPAR After the IPL of the IBM i the new DS8000 LUNs are recognized by IBM i 11 9 2 IBM i data migration 244 It is also possible to use native IBM i functions to migrate data from existing
160. al storage server IBM i 7 1 does not support logical volumes of size 8 59 as a load source unit LUN sizes IBM i can only use fixed logical volume sizes therefore configure more logical volumes than actual DDMs At a minimum use a 2 1 ratio For example with 146 GB DDMs use a maximum size of 70 56 GB LUNs When connecting DS8000 natively to IBM i it is important to consider using a smaller size of LUNs with IOP based adapters because they do not support SCSI command tag queuing Using smaller LUNs can reduce I O queues and wait times by allowing IBM i to support more parallel I Os In general 35 GB LUNs can be a good number to define with OP based adapters Unlike OP based adapters IOP less adapters support SCSI command tag queuing Therefore you can have bigger LUNs defined such as 70 56 GB LUNs or 141 1 GB LUNs Chapter 11 IBMiconsiderations 215 Connection of DS8000 series to IBM i with VIOS NPIV or VIOS VSCSI supports SCSI command tag queuing therefore use 70 56 GB LUNs or 141 1 GB LUNs with this type of connection Number of LUNs per port IOP based adapters can support up to 32 LUNs but for performance reasons do not use the maximum number Instead define less than 32 LUNs per adapter Table 11 6 on page 214 can help you to determine the number of LUNs per adapter because it shows the preferred capacity per adapter With multipath consider using double capacity per two adapters With native attachment using OP less
161. al partitions The current vir B e 9 e Client Fibre Channel Client Fibre Channel Client SCSI Server Serial Server Serial ATS_GERO_VIOS_MAIN_IP24 2 34 ATS_GERO_VIOS_MAIN_IP24 2 35 ATS_GERO_VIOS_MAIN_IP24 2 99 Any Partition Any Partition Slot Any Partition Any Partition Slot Total 5 Filtered 5 Selected 0 Figure 11 31 Assigning virtual Fibre Channel adapters Figure 11 32 shows both WWPNs for the virtual Fibre Channel adapter 14 in the Virtual Fibre Channel Adapter Properties window Virtual Fibre Channel adapter Adapter ID 14 Type of adapter Client Required True Server adapter ID 34 Virtual Fibre Channel Adapter Properties ATS_LPAR19_ITSO_JANA_IP42 WWPNs c050760035a80164 c050760035a80165 Server partition ATS_GERO_VIOS_MAIN_IP24 2 Figure 11 32 WWPNs of the virtual Fibre Channel adapter In the DS8000 the load source and other LUNs are assigned to both WWPN from each virtual Fibre Channel adapter Chapter 11 IBM i considerations 239 In VIOS the virtual Fibre Channel adapters from IBM i correspond to virtual devices vfchost12 and vfchost13 Figure 11 33 lsdev slots Slot Description Device s HEA 1 Logical I O Slot lhea0 ent0 U789D 001 DQD904G P1 T3 Logical I O Slot pci2 sissas0 U9117 MMA 06C6DE1 V2 C0 Virtual I O Slot vsa0 U9117 MMA 06C6DE1 V2 C2 Virtual I 0 Slot vasi0 U9117 MMA 06C6DE1 V2 C1l1 Virtual I 0 Slot
162. an be used to analyze PAV usage is available at the following website http www ibm com servers eserver zseries zos unix bpxalty2 html pavanalysis HyperPAV commands for setup control and status display Use the following commands for HyperPAV management and status information SETIOS HYPERPAV YES NO BASEONLY SET I0S xx D M DEV D I0S HYPERPAV DEVSERV QPAV dddd y vvvy In Example 12 8 the d m dev command provides system configuration information for the base address 0710 that belongs to an LSS with enabled HyperPAV Example 12 8 Display information for a base address in an LSS with enabled HyperPAV SY1 d m dev 0710 SY1 IEE1741I 23 35 49 DISPLAY M 835 DEVICE 0710 STATUS 0NLINE CHP 10 20 30 40 DEST LINK ADDRESS 10 20 30 40 PATH ONLINE Y Y Y Y CHP PHYSICALLY ONLINE Y Y Y Y PATH OPERATIONAL Y Y Y Y MANAGED N N N N CU NUMBER 0700 0700 0700 0700 MAXIMUM MANAGED CHPID S ALLOWED 0 DESTINATION CU LOGICAL ADDRESS 07 SCP CU ND 002107 000 1BM TC 03069A000007 OOFF SCP TOKEN NED 002107 900 1BM TC 03069A000007 0700 SCP DEVICE NED 002107 900 1BM TC 03069A000007 0710 HYPERPAV ALIASES IN POOL 4 In Example 12 9 address 0718 is an alias address belonging to a HyperPAV LSS If you see a HyperPAV alias in use or bound it is displayed as bound Example 12 9 Display information for an alias address belonging to a HyperPAV LSS SY1 D M DEV 0718 SY1 1IEE174I 23 39 07 DISPLAY M 838 DEVICE 0718 STATUS POOLED
163. ance to a well established code base POWER6 POWERS POWER6 DS8800 DS8000 DS8700 DS8000 DS8800 builds ona market proven reliable code base Figure 1 1 The DS8000 family of storage systems The focus of this book is on attaching hosts to the latest DS8000 storage system the DS8800 There are also several areas where the previous generation the DS8700 is mentioned because these systems share some common architecture elements Throughout this book if a feature or component is specific to a model the exact model name is used such as DS8700 or DS8800 If the information is of a more general nature then the less specific term DS8000 is used indicating that the information is relevant to all the models of the DS8000 family of products Important When information in this book applies to only a specific storage system model such as the DS8 amp 00 it is referenced explicitly References to DSS000 are more general and apply to the entire family of DS8000 storage systems not a specific model 2 IBM System Storage DS8000 Host Attachment and Interoperability 1 2 Client architecture overview To assist with upgrade planning or troubleshooting the client needs to provide current versions of the following items gt SAN A layout or map identifying all the components in the SAN along with contact information for storage administrators gt SAN A table that cross references LUNs from a hos
164. and Note the section at the end of Example 10 12 where the information about the logical and physical volumes is displayed Example 10 12 Displaying volume group details vgdisplay v vg01 Volume groups VG Name dev vg01 VG Write Access read write VG Status available Max LV 255 Cur LV 1 Open LV 1 Max PV 16 Cur PV 1 Act PV 1 Max PE per PV 8000 194 IBM System Storage DS8000 Host Attachment and Interoperability VGDA PE Size Mbytes Total PE Alloc PE Free PE Total PVG Total Spare PVs Total Spare PVs in use Logical volumes LV Name LV Status LV Size Mbytes Current LE Allocated PE Used PV Physical volumes PV Name PV Status Total PE Free PE Autoswitch 3071 3000 71 dev vg01 1vol1 available syncd 12000 3000 3000 1 dev disk disk14 available 3071 71 On Now the details regarding the physical volume s can be examined by using the pvdisplay command as shown in Example 10 13 Example 10 13 Displaying physical volume details pvdisplay dev disk disk14 Physical volumes PV Name VG Name PV Status Allocatable VGDA Cur LV PE Size Mbytes Total PE Free PE Allocated PE Stale PE I0 Timeout Seconds Autoswitch dev disk disk14 dev vg01 available yes 2 1 4 3071 71 3000 0 default On At this point it is time to increase the size of the DS8000 LUN After the change is made take the following steps at th
165. and Flash Update 153 8 4 Using the ATTO Configuration Tool 0 0 cette 153 BAe PALS oer ack asians AO a hn ak es a Aa dew gh ended Petes E EOE 154 8 4 2 Target Base ic feos At es SN to I eI tt es ae N 155 8 4 3 LUN Base once teat tag ye eo ae eos oe OE ea ae es eae Bee 155 B44 Pah ACtlons se ae tt pA EA tl E RA EEEN EEA ERA CY 155 8 4 5 Host bus adapter configuration 0 0 eee 156 8 5 Creating a file system on Apple Mac OS X 1 2 0 156 8 5 1 Using the GUI Disk Utility 2 2 00 0 0 157 8 5 2 Using the CLI diskutil 0 0 0 00 157 8 6 Troubleshooting saree fe teeth tg ete bs we eats aaa deena Gob mie eke wand AOA Sale eal 157 8 6 1 Useful Utilities on Apple Mac OSX 0 0000 157 8 6 2 Troubleshooting checklist 0 0 0 0 158 Chapter 9 Solaris considerations 00 000 ccc ee 159 Contents v vi 9 1 Working with Oracle Solaris and DS8000 0 0 0 ee 160 9 2 Locating the WWPNs of your HBAS 000 eect tees 160 9 3 Attaching Solaris to DS8000 1 2 eee 161 9 4 Multipathing in Solaris 2 0 0 tees 162 9 4 1 Working with IBM System Storage Multipath SDD 5 163 9 4 2 Using MPIO sa 2s vias tae iio BE ee A Bee a 164 9 4 3 Working with VERITAS Volume Manager dynamic multipathing 165 9 5 Expanding dynamic volume with VxVM and DMP 00 00 eee eee 167 9 6 Booting from SAN 1 2 0 t
166. and removing DS8000 volumes dynamically on Linux Unloading and reloading the Fibre Channel HBA adapter was the typical method to discover newly attached DS8000 volumes However this action is disruptive to all applications that use Fibre Channel attached disks on this particular host With recent Linux versions you can add newly attached LUNs without unloading the FC HBA driver To do this obtain the SCSI instances that your FC HBAs have then scan for new Fibre Channel attached devices by using the command syfs as shown in Example 6 37 Example 6 37 Scanning for new Fibre Channel attached devices x36501ab9 1s sys class fc_host host3 host4 x36501ab9 echo gt sys class scsi_host host3 scan X36501ab9 echo gt sys class scsi_host host4 scan The triple dashes represent the channel target and LUN combination to scan A dash causes a scan through all possible values A number will limit the scan to the given value New disk devices that are discovered with this method automatically get device nodes and are added to DM Multipath 106 IBM System Storage DS8000 Host Attachment and Interoperability Tip For certain older Linux versions you must force the FC HBA to perform a port login to recognize newly added devices by using the following command echo 1 gt sys class fc_host host lt ID gt issue_lip You need to run this command for all FC HBAs Removing a disk device or volume from Linu
167. anging the volume size Are you sure that you want to resize the volume y n y CMUCO0026I chfbvol FB volume 4791 successfully modified Rescan all disk devices or paths of a DS8000 volume The Linux SCSI layer picks up the new capacity when a rescan is initiated of each SCSI disk device or path through sysfs as shown in Example 6 47 Example 6 47 Rescan all disk devices or paths of a DS8000 volume x36501ab9 echo 1 gt sys class scsi_disk 3 0 0 3 device rescan x36501ab9 echo 1 gt sys class scsi_disk 3 0 1 3 device rescan x36501ab9 echo 1 gt sys class scsi_disk 4 0 0 3 device rescan X36501ab9 echo 1 gt sys class scsi_disk 4 0 1 3 device rescan Example 6 48 shows the message log that indicates the change in capacity Example 6 48 Linux message log indicating the capacity change of a SCSI device x36501ab9 tail var log messages Oct 13 16 52 25 x36501ab9 kernel 9927 105262 sd 3 0 0 3 sde 209715200 512 byte logical blocks 107 GB 100 GiB Oct 13 16 52 25 x36501lab9 kernel 9927 105902 sde detected capacity change from 53687091200 to 107374182400 Indicate the device change to DM Multipath using the resize_map command of multipathd as shown in Example 6 49 Afterwards you can see the updated capacity in the output of show topology Example 6 49 Resizing a multipath device x36501ab9 multipathd k resize map 3600507630af fc29f0000000000004791 ok x36501ab9 multipat
168. are called the StorEdge Traffic Manager Software STMS STMS is commonly known as multiplexed I O MPxIO in the industry and the remainder of this book references this technology as MPxIO gt IBM supports VERITAS Volume Manager VxVM Dynamic Multipathing DMP a part of the VERITAS Storage Foundation suite The multipathing technology that you use depends predominantly on your operating environment and your business requirements There are a few limitations depending on your operating system version your HBAs and whether you use clustering Details are available in the IBM System Storage DS8000 Host Systems Attachment Guide SC26 7917 One difference between the multipathing technologies is in whether they suppress the redundant paths to the storage MPxIO and DMP both suppress all paths to the storage except for one and the device appears to the application as a single path device However SDD allows the original paths to be seen but creates its own virtual device called a vpath for applications to use If you assign LUNs to a server before you install multipathing software you can see each LUN show up as two or more devices depending on how many paths you have In Example 9 4 the iostat nE command shows that the volume 75207814206 appears twice as c2t1d1 on the first HBA and as c3tid1 on the second HBA 162 IBM System Storage DS8000 Host Attachment and Interoperability Example 9 4 Device listing without multipath software
169. ase of a disk drive or SAN attached volume this name contains a string that uniquely identifies the volume This string ensures that every time this volume is attached to the system it gets the same name Therefore the issue where devices received separate names depending on the order they were discovered is no longer valid with newer Linux releases Multipathing Linux now has its own built in multipathing solution DM Multipath It is based on the device mapper a block device virtualization layer in the Linux kernel The device mapper is also used for other virtualization tasks such as the logical volume manager data encryption snapshots and software RAID DM Multipath overcomes the issues when only proprietary multipathing solutions existed gt The proprietary multipathing solutions were only supported for certain kernel versions Therefore systems followed the distributions update schedule gt Proprietary multipathing was often binary only and not supported by the Linux vendors because they were unable to debug the issues gt A mix of separate vendors storage systems on the same server or even separate types of the same vendor usually was not possible because the multipathing solutions cannot coexist Chapter 6 Linux considerations 85 Today DM Multipath is the only multipathing solution fully supported by both Red Hat and Novell for the enterprise Linux distributions It is available on all hardware platforms a
170. atform for Linux the Intel x86 32 or 64 bit architecture only allows direct mapping of DS8000 volumes to the host through Fibre Channel fabrics and HBAs The other platforms such as IBM System z and IBM Power Systems provide additional mapping methods to enable better exploitation of their much more advanced virtualization capabilities IBM Power Systems Linux running in a logical partition on an IBM Power System can get storage from a DS8000 either directly through an exclusively assigned Fibre Channel HBA or through a VIOS running on the system Direct attachment works the same way as documented previously with the other platforms However VIOS attachment requires specific considerations For other details about the way VIOS works and how it is configured see Chapter 4 Virtual I O Server considerations on page 47 For other publications that cover VIOS attachment on Power Systems see these books gt PowerVM Virtualization on IBM System p Introduction and Configuration Fourth Edition SG24 7940 gt IBM PowerVM Virtualization Managing and Monitoring SG24 7590 86 IBM System Storage DS8000 Host Attachment and Interoperability Virtual vscsi disks through VIOS Linux on IBM Power distributions contain a kernel module or driver for a virtual SCSI HBA which attaches the virtual disks provided by the VIOS to the Linux system This driver is called ibmvscsi Example 6 1 shows how the devices are viewed by the Linux system
171. ath being selected for IO will fail again a short time later due to intermittent link failure condition Furthermore the TIMEOUT error triggers FC recovery commands and the recovery commands also experience TIMEOUT error This TIMEOUT Recovery storm degrades I O performance severely In the worst case it might cause application failure A path not the last path will be set to FAILED state if I O timed out To recover a failed path due to TIMEOUT error the algorithm is different based on the timeout_policy setting gt retry_path The algorithm works the same way as previous version when a health check command succeeds a failed path due to TIMEOUT will be recovered immediately gt fail_path This setting requires two consecutive successful health check commands to recover a failed path due to TIMEOUT gt disable_path If a TIMEOUT failed path continuously experiences health check command TIMEOUT within a certain period of time the failed path due to TIMEOUT will be set to DISABLED OFFLINE state It will stay in the DISABLED state until user manually recovers it Tip The default setting of timeout_policy is fail_path 70 IBM System Storage DS8000 Host Attachment and Interoperability Considerations for fail_ over gt The fail_path and disable_path options are not supported under the fail_over algorithm Therefore the user who wants to change the device algorithm to fail_over must change
172. atile storage NVS size of that cluster z OS users will find that only the SETCACHE CFW ON OFF command is supported but other SETCACHE command options For example DEVICE SUBSYSTEM DFW NVS are not accepted The cache and NVS size reported by the LISTDATA command is somewhat less than the installed processor memory size The DS8000 licensed internal code uses part of the processor memory and it is not reported by LISTDATA Chapter 12 IBM System z considerations 253 12 3 6 System Management Facilities System Management Facilities SMF is a component of IBM Multiple Virtual Storage IBM MVS Enterprise Systems Architecture ESA System Product SP that collects I O statistics provided at the data set and storage class levels It helps monitor the performance of the direct access storage subsystem SMF collects disconnect time statistics that are summarized and reported at the data set level To support Solid State Drives SSD SMF is enhanced to separate DISC time for READ operations from WRITE operations Here are the two subtypes gt SMF 42 Subtype 6 This subtype records DASD data set level I O statistics I O response and service time components are recorded in multiples of 128 microseconds for the data set I O statistics section S42DSRDD is the average disconnect time for reads S42DSRDT is the total number of read operations gt SMF 74 Subtype 5 The DS8000 provides the ability to obtain cache stati
173. ays of describing the mass storage the legacy view and the agile view The legacy view of storage devices is simply the legacy DSFs and legacy hardware paths utilizing the familiar description of cXtYdZ that was in common use for a long time The agile view uses LUN hardware paths and persistent DSFs Different commands offer their information in the different views Some commands such as ioscan will show the legacy view by default But they can be made to show the agile view with the use of the N option Be aware of this option as you work through the remainder of this chapter 10 1 2 Notes on multipathing For providing a fault tolerant connection to a HP UX system there are three possible multipathing solutions However only one native multipathing is appropriate for the DS8000 A more complete discussion of multipathing can be found in Multipathing on page 187 The following options are available gt HP UX native multipathing introduced with HP UX 1 1iv3 gt HP PVLINKS as provided in earlier releases of HP UX PVLINKS is still available in HP UX 11iv3 However when attaching to DS8000 you need to use native multipathing gt IBM Multipath Subsystem Device Driver SDD HP UX 11iv3 SDD from IBM is not supported with HP UX 11iv3 Native multipathing from HP offers load balancing over the available I O paths and I O path failover in case of a connection failure The older PVLINKS provides a failover solution but do
174. bility 0 2 0 2 0 2 0 2 Manage Profiles r anage Custom Groups Save Current Configuration Figure 11 27 HMC Manage Profiles 2 In the Manage Profiles window select the IBM i LPAR expand the Actions menu and select Edit as shown in Figure 11 28 ter assignments for a profile by editing the profile ontains the resource configuration for the partition You can modify the processor Delete Select Action m Activate ctv atus M default Default Profile Last Activated Close Help javascript menultemLaunchaActionQ 236 IBM System Storage DS8000 Host Attachment and Interoperability Figure 11 28 HMC Edit profile 3 Select the table Tagged I O and expand the pull down menu under Load source Figure 11 29 Select the adapter that connects the boot from SAN volume This figure shows an example of tagging an 8 GB IOP less Fibre Channel adapter through the load source that is connected x3550 lab 8 Manage Profiles Mozilla Firefox IBM Edition he I ole tan ELL aya k https 9 155 57 94 hmc wel T2d02c Logical Partition Profile Properties default ATS_LPAR19_ITSO_JANA_IP42 p6 570 2 O6C6DE1 ATS_LPAR19_ITSO_JANA_IP42 Logical Virtual Power Host General Processors Memory I O Adapters Controlling Settings prema 1 0 OptiConr LHEA Tagged I O devices for this partition profile are detailed below
175. ble Virtual SCSI Disk Drive Iscfg vl hdisk2 hdiskO U9117 MMB 100DBCP V9 C31 T1 L8100000000000000 Virtual SCSI Disk Drive Isattr El hdisk2 PCM PCM friend vscsi Path Control Module False algorithm fail_over Algorithm True hcheck_cmd test_unit_rdy Health Check Command True hcheck_interval 20 Health Check Interval True hcheck_mode nonactive Health Check Mode True max_transfer 0x40000 Maximum TRANSFER Size True pvid 00f60dbc823d03050000000000000000 Physical volume identifier False queue_depth 3 Queue DEPTH True reserve policy no_reserve Reserve Policy True However the UDID of the virtual SCSI disk devices can be used to properly identify which DS8000 LUN is used on the VIOS as backing device The UDID is stored in the AIX Object Data Manager ODM and can be listed using the odgmet q attribute unique_id CuAt command as shown in Example 5 6 Example 5 6 Listing the UDID for DS8000 backed Virtual SCSI Disk devices in AIX client partition odmget q attribute unique_id CuAt CuAt name hdiskO attribute unique_id value 3520200B75TV181900407210790003IBMfcpO5VDASDO3AIXvscsi type R generic rep n nls_index 0 The UDID in Example 5 6 contains the original UDID of the DS8000 LUN as shown on the VIOS using the chkdev command in Example 4 9 on page 55 In Example 5 6 the UDID value of 3520200B75TV181900407210790003IBMfcp05VDASDO3AIXvscsi shown for hdisk0 contains the serial number for the DS8000 storage image 75T
176. bout This Mac More Info ATTO Configuration Tool This tool assists in identifying a failed or inactive path kextstat cmd Use the kexstat command to check which ATTO FC driver is loaded into memory Run it in a terminal window as shown in Example 8 1 on page 153 8 6 2 Troubleshooting checklist 158 The following items can be helpful if you need to report a problem to ATTO or IBM This list is not exhaustive but it helps to have this information handy when troubleshooting any adapter or connectivity problem gt Computer Model see Apple System Profiler Hardware Operating System see Apple System Profiler Software OS Patch Level see Apple System Profiler Software PCI or PCle slot and type see Apple System Profiler Hardware PCI Cards ATTO driver version from the Configuration Tool see ATTO Configuration Tool localhost Celerity FC xxx Basic Info List of all of the devices attached to the ATTO HBA see ATTO Configuration Tool or Apple System Profiler gt Hardware Fibre Channel select SCSI Target Device Did this configuration ever work Is this a new error that just started happening or is this an error that was around since the first use Can the error be duplicated Does the error occur sporadically randomly or can it be reproduced each and every time Apple System Profiler Output ASP ioreg output open terminal window to run command and redir
177. c7t2d1s2 Ethan02 Ethan ENABLED c6tld2s2 c7t2d2s2 Ethan03 Ethan ENABLED c7t2d2s2 c7t2d2s2 Ethan03 Ethan ENABLED c6t1ld3s2 c7t2d3s2 Ethan04 Ethan ENABLED c7t2d3s2 c7t2d3s2 Ethan04 Ethan ENABLED Now you create volumes using the device name listed under the DANAME column In Figure 9 1 a volume is created using four disks even though there are actually eight paths File Tools Actions Window Help gt Se a s Es Connect Disconnect New volume New Group Search ft System Management Console 9 155 66 18 fgg Controllers 831 Disk Groups Disks E citOdds2 3 citidds2 S amp S cit2d0s2 E cit3d0s2 E cit4d0s2 S amp S cit5d0s2 e c6tid0s2 SS c7t2dis2 E c7t2d2s2 i E c7t2d3s2 E ga Enclosures E Normal usage High usage W Critical usage met a Device Status Size 6 citOdds2 Not Initialized Not Initialized Tea Not Initialized 5 cit3d0s2 Not Initialized I cit4dOs2 Not Initialized Ethan Ethan Ethan Ethan mare Name Target Host Time started Status Progress 2 Create Disk Group c6tidOs2 9 155 66 18 November 2 Successful Figure 9 1 VERITAS DMP disk view 3 9 155 66 18 BE Tuning parameters As with other multipathing software DMP provides a number of parameters that you can tune to maximize the performance and availability in your environment For example it is possible to set a loa
178. cal volume device Example 6 54 Example 6 54 Mounting the source volume x36501ab9 mount dev mapper vg_ds8000 1lv_itso mnt lv_itso x36501ab9 mount dev mapper vg_ds8000 1lv_itso on mnt lv_itso type ext3 rw 2 Make sure the data on the source volume is consistent for example by running the sync command 3 Create the FlashCopy on the DS8000 and map the target volumes 46a0 and 47a0 to the Linux host 4 Initiate a device scan on the Linux host see Adding and removing DS8000 volumes dynamically on Linux on page 106 for details DM Multipath automatically integrates the FlashCopy targets as shown in Example 6 55 Example 6 55 Checking DM Multipath topology for target volume x36501ab9 multipathd k show topology 3600507630af fc29f00000000000046a0 dm 6 IBM 2107900 size 50G features 1 queue_if_no_path hwhandler 0 wp rw policy round robin 0 prio 1 status active 3 0 0 4 sdf 8 80 active ready running 3 0 1 4 sd 8 176 active ready running 4 0 0 4 sdr 65 16 active ready running 4 0 1 4 sdx 65 112 active ready running 3600507630af fc29f00000000000047a0 dm 7 IBM 2107900 size 50G features 1 queue_if_no_path hwhandler 0 wp rw policy round robin 0 prio 1 status active 3 0 0 5 sdg 8 96 active ready running 1 5 sdm 8 192 active undef running 5 sdy 65 128 active ready running 3 0 4 0 1 4 0 0 5 sds 65 32 active ready running Chapter 6 Linux considerations 113 I
179. can find at the following website http support qlogic com support oem_ibm asp Emulex Corporation You can find the Emulex home page at this website http www emulex com Emulex also has a page with content specific to IBM storage systems at the following website http www emulex com products host bus adapters ibm branded html Oracle Oracle ships its own HBAs which are Emulex and QLogic based However the native HBAs from Emulex and Qlogic can be used to attach servers running Oracle Solaris to disk systems In fact these native HBAs can even be used to run StorEdge Traffic Manager software For more information about the Oracle HBAs see the following websites gt For Emulex http www oracle com technetwork server storage solaris overview emulex corpor ation 136533 html Chapter 2 Open systems considerations 7 gt For QLogic http www oracle com technetwork server storage solaris overview qlogic corp 139073 htm Hewlett Packard HP ships its own HBAs which are Emulex and QLogic based gt Emulex publishes a cross reference at the following website http www emulex hp com interop matrix index jsp mfgId 26 gt QLogic publishes a cross reference at the following website http driverdownloads qlogic com QLogicDriverDownloads UI Product_detail aspx oemid 21 Atto Technology Inc Atto Technology supplies HBAs for Apple Macintosh which is supported by IBM for the attachment of the DS8000 The Att
180. ce mapper multipathing 98 To gain redundancy and optimize performance connect a server to a storage system through more than one HBA fabric and storage port This server connection results in multiple paths from the server to each attached volume Linux detects such volumes more than once and creates a device node for every instance To use the path redundancy and increased I O bandwidth and at the same time maintain data integrity you need an additional layer in the Linux storage stack to recombine the multiple disk instances into one device Linux provides its own native multipathing solution It is based on the device mapper a block device virtualization layer in the Linux kernel Therefore it is called DM Multipath Figure 6 2 illustrates how DM Multipath is integrated into the Linux storage stack Figure 6 2 Device mapper multipathing in the Linux storage stack IBM System Storage DS8000 Host Attachment and Interoperability The device mapper is also used for other virtualization tasks such as the logical volume manager data encryption snapshots and software RAID DM Multipath is able to manage path failover failoack and load balancing for various separate storage architectures In simplified terms DM Multipath consists of four main components gt The dm multipath kernel module takes the I O requests that go to the multipath device and passes them to the individual devices representing the paths gt The mul
181. chapter provides information for the specifics of attaching IBM System Storage DS8000 systems to host systems running Hewlett Packard UNIX HP UX There are multiple versions of HP UX available however this chapter presents information based on HP UX 11iv3 The following topics are covered Working with HP UX Available resources Identifying available HBAs Identifying WWPNs of HBAs Configuring the HP UX host for the DS8000 Multipathing Working with VERITAS Volume Manager on HP UX Working with LUNs vvvvvvvyy Copyright IBM Corp 2011 2013 All rights reserved 181 10 1 Working with HP UX This chapter presents configuration information for attaching a host running HP UX to the point where the host is capable of running I O to the DS8000 device HP UX is the UNIX System V implementation offered by Hewlett Packard HP The DS8000 supports the attachment to hosts running HP UX 11i or later There is a release name and identifier assigned to each release of HP UX that can be displayed with the uname r command The response of B 11 31 indicates that the host is running HP UX 11i v3 as shown in Example 10 1 Example 10 1 Determining the version of HP UX uname r B 11 31 10 1 1 The agile view HP UX commands that deal with storage have a choice in how the information is presented The industry usage of Device Special Files DSF discussed further in 10 5 1 Device special files on page 185 leads to two different w
182. ck disable auto failback enable For enabling MPxIO support for 3rd party symmetric device need an entry similar to following in this file Just replace the SUN SENA part with the Vendor ID Product ID for the device exactly as reported by Inquiry cmd device type scsi options list device type scsi options list SUN SENA symmetric option IBM 2109700 symmetric option Chapter 9 Solaris considerations 177 symmetric option 0x1000000 symmetric option 0x1000000 4 Enter the stmsboot e command to reactivate MPxIO as shown in Example 9 27 Example 9 27 Reactivating MPxIO bash 3 00 stmsboot e WARNING stmsboot operates on each supported multipath capable controller detected in a host In your system these controllers are devices pci 80 2000 em1 x 2 fpe0 0 devices pci 80 2000 em1x 2 fp 1 0 devices pci 80 2000 em1x 1 fpe0 0 devices pci 80 2000 em1x 1 fp 1 0 If you do NOT wish to operate on these controllers please quit stmsboot and re invoke with D fp mpt to specify which controllers you wish to modify your multipathing configuration for Do you wish to continue y n default y Checking mpxio status for driver fp Checking mpxio status for driver mpt WARNING This operation will require a reboot Do you want to continue y n default y The changes will come into effect after rebooting the system Reboot the system now y n default y updating p
183. ck is performed The attribute hcheck_mode is set to the preferred value of nonactive by default which means that health check commands are sent down all paths that have no active I O including paths with a state of failed In that case after one VIOS fails and then comes back manually set the path as available Changing hcheck_interval is advisable for all virtual SCSI hdisk devices as shown in Example 5 9 Example 5 9 Changing the hcheck_interval to 60 seconds for a virtual SCSI disk chdev 1 hdisk2 a hcheck_interval 60 P hdisk2 changed Chapter 5 AIX considerations 67 The hcheck_interval must never be set lower than the read or write time out value of the underlying physical disk on the VIOS which is 60 seconds for a DS8000 volume being used as backing device with SDDPCM see Example 4 10 on page 56 Otherwise if the Fibre Channel adapter has an error there will be new health check requests sent before the running ones time out and it can lead to an adapter failure condition The hcheck_interval attribute Set the hcheck_interval attribute value to 60 Also as shown in Example 5 5 on page 65 the default queue depth of a virtual SCSI disk device is set to 3 When a DS8000 volume is used as backing device on the VIOS with a default queue depth of 20 as shown in Example 4 10 on page 56 with SDDPCM consider increasing the queue depth of the virtual SCSI disk device to the same value shown in Example 5 10 for best performance
184. ck size and capacity in blocks of the device gt sginfo dev sgx prints SCSI inquiry and mode page data It also helps you manipulate the mode pages 6 5 4 Booting Linux from DS8000 volumes You can configure a system to load the Linux kernel and operating system from a SAN attached DS8000 volume boot from SAN This section explains this process using SLES11 SP1 on an x86 server with QLogic FC HBAs The Linux boot process In order to understand the configuration steps required to boot a Linux system from SAN attached DS8000 volumes you need a basic understanding of the Linux boot process Therefore this process is briefly summarized in the following sections for a Linux system until the point in the process where the well known login prompt is presented OS loader The system firmware provides functions for rudimentary I O operations for example the BIOS of x86 servers When a system is turned on it first performs the power on self test POST to check the hardware available and verify that everything is working Then it runs the operating system loader OS loader which uses basic I O routines to read a specific location on the defined system disk and executes the code it contains This code either is part of the boot loader of the operating system or it branches to the location where the boot loader resides 116 IBM System Storage DS8000 Host Attachment and Interoperability If you want to boot from a SAN attached disk ensure
185. conflicts More information can be found in the following website http technet microsoft com en us 1ibrary cc739757 28v ws 10 29 aspx 3 3 2 SDD support SDD handles path reclamation in a Windows cluster environment subtly different from a non clustering environment When the Windows server loses a path in a non clustering environment the path condition changes from open to dead and the adapter condition changes from active to degraded The adapter and path condition will not change until the path is made operational again When the Windows server loses a path in a clustering environment the path condition changes from open to dead and the adapter condition changes from active to degraded However after a period of time the path condition changes back to open and the adapter condition changes back to normal even if the path has not been made operational again Tip The adapter status changes to a degraded state when paths are active and to a failed state when there are no active paths The datapath set adapter offline command operates separately in a clustering environment as compared to a non clustering environment In a cluster environment the datapath set adapter offline command does not change the condition of the path if the path is active or being reserved 3 4 Using Multipath Input Output for Windows 2003 and 2008 Microsoft s Multipath Input Output MPIO solutions work in conjunction with device specific modules DSMs
186. considerations apply regarding Virtual Storage Extended Enterprise Systems Architecture VSE ESA and z VSE support 274 gt An APAR is required for VSE 2 7 to exploit large volume support but support for EAV is still not provided VSE has a default MIH timer value to 180 seconds You can change this setting to the suggested DS8000 value of 30 seconds by using the SIR MIH command which is documented in Hints and Tips for VSE ESA 2 7 at this website http www ibm com servers eserver zseries zvse documentation IBM System Storage DS8000 Host Attachment and Interoperability 12 8 I O Priority Manager for z OS DS8000 I O Priority Manager is a licensed function feature introduced for IBM System DS8000 storage systems with DS8000 Licensed Machine Code LMC R6 1 or higher Starting from release 6 2 LMC level 7 6 20 xx I O priority is also working on z OS environments and it interacts with Workload Manager eWLM It enables more effective storage consolidation and performance management and the ability to align quality of service QoS levels to separate workloads in the system that are competing for the same shared and possibly constrained storage resources DS8000 I O Priority Manager constantly monitors system resources to help applications meet their performance targets automatically without operator intervention The DS8000 storage hardware resources that are monitored by the I O Priority Manager for possible contention are t
187. controller PCI Express gt 5273 PCle LP 8 Gb 2 Port Fibre Channel adapter gt 5729 PCle2 8 Gb 4 port Fibre Channel adapter For Fibre Channel over Ethernet connection gt 5708 10 Gb FCoE PCle Dual Port adapter gt 5270 PCle LP 10 Gb FCoE 2 port adapter NPIV connection requires an NPIV enabled SAN switches for connection of DS8000 Connecting DS8000 series to an IBM i through VIOS NPIV in IBM BladeCenter requires one of the following adapters gt 8271 QLogic 8 GB Fibre Channel Expansion Card CFFh gt 8242 QLogic 8 GB Fibre Channel Card ClOv gt 8240 Emulex 8 GB Fibre Channel Expansion Card ClOv For Fibre Channel over Ethernet connection gt 8275 QLogic 2 port 10 Gb Converged Network Adapter CFFh With VIOS NPIV up to 64 LUNs can be assigned to a port in a virtual Fibre Channel adapter in IBM i The IBM i with virtual Fibre Channel adapters used with VIOS NPIV provides SCSI command tag queuing support of DS8800 system with the queue depth being 6 I O operations 11 2 3 Attachment with VIOS VSCSI Connecting DS8000 series to an IBM i with VIOS virtual SCSI in POWER servers requires one of the following adapters in VIOS gt 5774 4 GB Fibre Channel Disk Controller PCI Express gt 5735 8 GB Fibre Channel Disk Controller PCI Express Attaching DS8000 to an IBM i through VIOS VSCSI in BladeCenter requires one of the following adapters gt Any of the 8 GB adapters and FCoE adapter listed under the VIOS NPI
188. ction ACS routines Copy data at the volume level with either IBM TDMF DFSMSdss Peer to Peer Remote Copy PPRC DFSMS Copy Services Global Mirror Metro Mirror Global Copy or FlashCopy Copy data at the data set level with SMS attrition ZDMF z OS Dataset Mobility Facility DFSMSdss and DFSMShsm Good volume candidates for EAV include DB2 zFS CICS VSAM record level sharing RLS and Information Management System IMS VSAM Poor EAV candidates are DFSMShsm migration level 1 Backup or migration level 2 although small data set packaging is eligible for cylinder managed space Other poor EAV candidates are work volumes TSO generation data group batch load libraries and system volumes DFSMSdss and DFSMShsm considerations when migrating to EAV are described in the Redbooks publication DFSMS V1 10 and EAV Technical Guide SG24 7617 12 5 FICON specifics for a z OS environment With FC adapters that are configured for FICON the DS8000 series provides the following configuration capabilities Either fabric or point to point topologies A maximum of 128 host adapter ports depending on the DS8800 processor feature A maximum of 509 logins per Fibre Channel port A maximum of 8192 logins per storage unit A maximum of 1280 logical paths on each Fibre Channel port Access to all control unit images over each FICON port A maximum of 512 logical paths per control unit image vvvvvvy 12 5 1 Overview FICON host c
189. d then a LUNs controlled by that Target will be affected by configuration changes There are three options for setting the method used for Load Balancing across multiple paths to the storage The Load Balancing policies are as follows gt Pressure The path with the fewest bytes in the queue will be used if needed gt Queue Depth The path with the fewest commands in the queue will be used gt Round Robin The least used path is selected By clicking Next you can modify the highlighted path for write and read mode Save Delete Configurations This option saves the configuration and will apply any changes made If new LUNs were created then you need to save the new configuration otherwise the newly created LUN s are not visible Locate Use this button to blink the particular FC adapter s channel LED Reset Statistics Statistics are set to 0 no data is preserved when this button is pressed Refresh If the connection status for a path changes the Configuration Tool will automatically refresh the display The Configuration Tool will periodically 30 seconds refresh the path statistics for all devices Nevertheless you can manually refresh the displayed information by choosing refresh 8 4 5 Host bus adapter configuration Changes to the adapter configuration are made in the NVRAM tab of the ATTO Configuration Tool In the Device Listing section select the adapter channel then select NVRAM at the top of the
190. d Volume CSV 00 0000 cece ee 44 3 10 6 Best practices anf deans a s Het cet ear ieee SA aries oe eo aries Bee 45 Chapter 4 Virtual I O Server considerations 0 0000 c eee ene eeee 47 4 1 Working with IBM Virtual I O Server 000 0000 eee 48 4 2 Using VSCSI with IBM VIOS and DS8000 2 2 2 2 anaana 49 4 3 Using NPIV with IBM VIOS and DS8000 02 02 ee 58 Chapter 5 AIX considerations 00 000 c cee 61 5 1 Attaching native Fibre Channel 0 00 c cette eee 62 5 1 1 ASSIGNING VOIUMES oa bn uyeee Copa tte ee decided ya Poway ea a ate AION A 62 5 1 2 Using node port ID virtualization NPIV oaaae 0 0002 c eee eee 64 5 2 Attaching virtual SCS 0 20 00 00 cette eee 65 5 3 Important additional considerations 0000 cee eee 68 5 3 1 Queue depth tuning 0 0 tees 69 5 3 2 timeout_policy attribute sae a eee 70 5 3 3 max xfer size attribute 2 0 0 eee 71 5 3 4 Storage unit with multiple IBM Power Systems hosts running AIX 71 5 4 Multipathing with AIX 0 eet ee 72 54 1 SDD for AlX eri aaa tects Walls Gente bene ele Por nies a eee aed 72 5 4 2 SDDPCM for AIX matien a a i e eee eee 74 5 5 Configuring LVM iieiea iaeei ia m a y teeta 76 DST LVM Striping 2 5 fri re adel Se oe ha tie a Sa A Pe 76 5 5 2 Inter physical volume allocation policy 0 0 00 cee eee 77 25 32 EVM MNO eisai ase ETE op Ble oth ee ne Gf Queene nip Go
191. d balancing policy to dictate how the I O be shared between the separate paths It is also possible to select which paths get used and in which order in case of a failure You can find further details about the capabilities of DMP at the following website http www symantec com business support index page content amp i d HOWT074584 166 IBM System Storage DS8000 Host Attachment and Interoperability 9 5 Expanding dynamic volume with VxVM and DMP Starting with licensed machine code 5 3 xx xx it is possible to expand a volume in the DS8000 even if it is mapped to a host However a volume that is in a FlashCopy Metro Mirror or Global Mirror relationship cannot be expanded unless the relationship is removed which means the FlashCopy Metro Mirror or Global Mirror on that volume must be removed before you can expand the volume Information in this section explains how to expand a volume on a Solaris host with VxVM on a DS8000 volume To display the volume size use the command 1sfbvol as shown in Example 9 9 Example 9 9 Isfbvol fullid before volume expansion dscli gt Isfbvol fullid 4704 Name ID accstate datastate configstate deviceMIM datatype extpool cap 2 30B cap 10 9B cap blocks ITSO_v880_4704 IBM 2107 7520781 4704 Online Normal Normal 2107 900 FB 512 IBM 2107 7520781 P53 12 0 25165824 In Example 9 9 the capacity is 12 GB and the volume ID is 4704 To determine the corresponding disk on the Solaris host install t
192. d by connecting DS8000 to IBM i with multiple VIOS each path through one VIOS Multipath with VIOS in NPIV or VSCSI requires VIOS level V2 1 2 or higher Console management levels Make sure that the Hardware Management Console HMC of POWER server is on the latest firmware level or that the Systems Director Management Console SDMC of POWER server or BladeCenter is on the latest software level 11 1 3 Overview of hardware and software requirements This section presents two tables that list hardware and software combinations for the connection methods The tables provide a quick overview of the possible combinations that are currently supported For more detailed up to date information see the websites listed in section 11 1 4 Useful websites on page 204 Table 11 1 shows an overview of supported POWER and Blade models and IBM i software levels Table 11 1 Supported hardware and software levels DS8800 attach Server i5 4 i6 1 i7 1 Native POWER7 Yes 4 Yes VIOS NPIV VIOS VSCSI VIOS NPIV Blade servers Yes 2 gt Yes gt VIOS VSCSI based on POWER7 Native IBM Yes 2 Yes VIOS NPIV POWER6 VIOS VSCSI Chapter 11 IBM i considerations 203 11 1 4 Useful websites 204 DS8800 attach Server i5 4 i6 1 i7 1 VIOS NPIV Blade servers Yes 25 Yes VIOS VSCSI based on POWER6 Native POWER6 Yes Yes Yes VIOS NPIV POWER6 Yes Yes VIOS VSCSI Native POWER5 Yes Yes Yes POW
193. d disk the standard initRAMFS must be extended with the FC HBA driver and the multipathing software In recent Linux distributions it is done automatically by the tools that create the initRAMFS image After the root file system is accessible the kernel starts the init process The init process The init process brings up the operating system itself such as networking services user interfaces and so on At this point the hardware is already completely abstracted Therefore init is not platform dependent and there are not any SAN boot specifics A detailed description of the Linux boot process for x86 based systems can be found on IBM developerWorks at this website http www ibm com devel operworks 1inux 1ibrary 1 1inuxboot 6 5 5 Configuring the QLogic BIOS to boot from a DS8000 volume To configure the HBA to load a BIOS extension that provides the basic I O capabilities for a SAN attached disk see 2 2 Using the DS8000 as a boot device on page 8 Tip For IBM branded QLogic HBAs the BIOS setup utility is called FAST Util Chapter 6 Linux considerations 117 6 5 6 Understanding the OS loader considerations for other platforms The BIOS is the method that an x86 system uses to start loading the operating system You can find information and considerations on how it is done on other supported platforms in this section IBM Power Systems When you install Linux on an IBM Power System server or LPAR the Lin
194. determine the necessary number of physical Fibre Channel adapters in VIOS The guidelines are obtained by modeling port utilization with Disk Magic for the DS8700 connected to an open server with 70 reads and 50 reads and with 16 KB transfer sizes and 120 KB transfer sizes The maximum amount of IOPS to keep port utilization under 60 is modeled Table 11 7 shows the maximum amount of IOPS and MBps at separate read and write ratios and separate transfer sizes to keep port utilization in VIOS under the preferred 60 Table 11 7 Sizing the Fibre Channel adapters in VIOS Percentage of Transfer size Maximum IOPS per one port Maximum MBps per one reads per blocksize in VIOS port in VIOS second 8 GB 4GB 8 GB 4 GB adapter adapter adapter adapter 70 16 18 500 289 70 120 2 900 340 50 16 26 000 406 50 120 3 900 457 Port in an IOP less adapter Avoid using one port in an IOP less adapter to connect disk storage and the other port to connect to tape 214 IBM System Storage DS8000 Host Attachment and Interoperability 11 4 Sizing and numbering of LUNs This section provides information about the sizes of DS8000 LUNs for IBM i which sizes of LUNs to use and how many LUNs to assign to a port to help achieve good performance 11 4 1 Logical volume sizes IBM i connected natively or with VIOS NPIV is supported on DS8000 as fixed block FB storage Unlike other open systems using the FB archi
195. disks attached to this controller will be removed but not deleted Figure 3 25 Add hard drive Chapter 3 Windows considerations 43 4 Select Physical hard disk and select the disk from the drop down list then click OK Ca Hard Drive You can change how this virtual hard disk is attached to the virtual machine IF an operating system is installed on this disk changing the attachment might prevent the virtual machine from starting Controller Location scsi Controller 7 fo in use x Media You can compact or convert a virtual hard disk by editing the vhd file Specify the Full path to the file virtual hard disk vhd file Physical hard disk Disk 3 20 00 GB Bus 0 Lun 2 Target 0 x If the physical hard disk you want to use is not listed make sure that the disk is offline Use Disk Management on the physical computer to manage physical hard disks Figure 3 26 Add physical hard disk Tip The disk to be used as a pass through disk can contain a file system and data when being attached This method can be used to move data between different servers both physical and virtual 3 10 5 Cluster Shared Volume CSV Microsoft Server 2008 R2 Hyper V provides a functionality called Cluster Shared Volume which allows concurrent access of multiple Hyper V hosts to the same NTFS volume To use CSV the following prerequisites need to be fulfilled gt All hosts involved are confi
196. do not depend on the disk capacity for example a rank of 300 GB 15 K RPM Fibre Channel DDMs supports the same maximal number of disk operations per second as a rank of 146 GB 15 K RPM Fibre Channel DDMs The following measured maximum disk operations per second are for other DDM types One 6 P RAID 5 rank of 15 K RPM Fibre Channel disks in DS8100 DS8300 and DS8700 can handle maximum 1800 disk operations per second for 100 DDM utilization One 6 P RAID 5 rank of 15 K RPM SAS disks in DS8800 can support a maximum of 2047 disk operations per second for 100 DDM utilization One 6 P RAID 5 rank of large form factor LFF SSD in DS8100 DS8300 and DS8700 is capable of supporting a maximum of 25 200 disk operations per second at 100 SSD utilization One 6 P RAID 5 rank of small form factor SFF SSD in DS8800 can a support maximum of 43 000 disk operations per second at 100 disk SSD utilization v y v v Based on these values it is possible to calculate how many host IOPS each rank can handle at the preferred utilization of 40 for IOP based adapters or 60 for IOP less adapters Examples of maximal host I Os per hard disk drive HDD rank for the workloads with 70 reads per second and 50 reads per second as shown in Table 11 4 This calculation is done for an average of maximal supported disk operations in a rank with spares and a rank without spares It calculates 20 of read cache hit and 30 of write cache efficiency T
197. e 11 d When BIOS configurations are complete leave the utility and reboot the server 12 To exit the configuration for the HBA press ESC twice 13 When you see the Configuration settings modified message in the lower panel shown in Figure 2 6 select Save changes default and press Enter electable Boot Settings Selectable Boot Enabled Primary Boot Port Nane Lun Tini AAAAAAAAAAAAAAG A onfiguration Settings Adapter Settings i tahla Ront Gott inne Configuration settings modified Do not save changes Figure 2 6 Checking and saving changes Configuring the BIOS for both HBAs gives you the advantage of a redundant connection to the boot device and enables the system to start even if part of the connectivity for example a switch is unavailable Important Redundant BIOS connections to the boot volume enable the system to start even if a path to the storage system is missing A failover capability is not provided when a path fails during the boot process Between the start of the boot process and the activation of the operating system s multipathing the system is not protected against a path failure 12 IBM System Storage DS8000 Host Attachment and Interoperability If you see messages such as the messages in Figure 2 7 you have configured the BIOS correctly The QLogic BIOS is enabled and the system attempts to start from the defined DS8000 volume Drive Letter C is Moved to Drive Lette
198. e HP UX operating system level 1 Unmount the file system that has 1vol1 mounted 2 Deactivate volume group vg00 vgchange a n vg00 Chapter 10 HP UX considerations 195 3 Use the vgmodify command to let HP UX know that the LUN size was altered as shown in Example 10 14 Example 10 14 Applying the vgmodify command vgmodify dev vg01 Current Volume Group settings Max LV 255 Max PV 16 Max PE per PV 8000 PE Size Mbytes 4 VGRA Size Kbytes 1088 dev rdisk disk14 size changed from 12582912 to 18874368kb An update to the Volume Group IS required New Volume Group settings Max LV 255 Max PV 16 Max PE per PV 8000 PE Size Mbytes 4 VGRA Size Kbytes 1088 New Volume Group configuration for dev vg01 has been saved in etc lvmconf vg01 conf Old Volume Group configuration for dev vg01 has been saved in etc lvmconf vg01 conf old Starting the modification by writing to all Physical Volumes Applying the configuration to all Physical Volumes from etc 1lvmconf vg01 conf Completed the modification process New Volume Group configuration for dev vg01 has been saved in etc lvmconf vg01 conf old Volume group dev vg01 has been successfully changed 4 Reactivate volume group vg01 vgchange a y vg01 Run the pvdisplay command again Example 10 15 reflects the new physical volume details indicating that HP UX recognizes the LUN resizing Example 10 15 inquiring the physical volume pvdispla
199. e SSIC you can use the Storage Customer Opportunity REquest SCORE process formerly known as Request for Price Quotations RPQ Contact your IBM storage sales specialist or IBM Business Partner for submission of a SCORE Initiating the process does not guarantee that the desired configuration will be supported Any approval depends on the technical feasibility and the required test effort A configuration that equals or is similar to one of the already approved configurations is more likely to become approved than a completely separate configuration Chapter 2 Open systems considerations 13 2 4 Multipathing support for Subsystem Device Drivers To ensure maximum availability most clients choose to connect their open systems hosts through more than one Fibre Channel path to their storage systems An intelligent SAN layout helps protect from failures of Fibre Channel HBAs SAN components and host ports in the storage system Certain operating systems however cannot work natively with multiple paths to a single disk putting the data s integrity at risk The risk occurs because multiple write requests can be issued to the same data and nothing takes care of the correct order of writes To use the redundancy and increased input output I O bandwidth when using multiple paths an additional layer is needed in the operating system s disk device representation concept This additional layer recombines the multiple disks seen by the HBAs
200. e ZDAC discovery IBM System Storage DS8000 Host Attachment and Interoperability 12 3 4 Performance statistics Because a logical volume is no longer allocated on a single RAID rank or single device adapter pair the performance data is now provided with a set of rank performance statistics and extent pool statistics The RAID rank reports are no longer reported by IBM Resource Measurement Facility IBM RMF and IDCAMS LISTDATA batch reports RMF and IDCAMS LISTDATA are enhanced to report logical volume statistics that are provided on the DS8000 These reports consist of back end counters that capture the activity between the cache and the ranks in the DS8000 for each individual logical volume These rank and extent pool statistics are disk system wide instead of volume wide only 12 3 5 Resource Measurement Facility The Resource Measurement Facility RMF supports DS8000 from z OS V1 4 and later with APAR number OA06476 and PTFs UA90079 and UA90080 RMF was enhanced to provide monitor and III support for the DS8000 storage system Statistics The Disk Systems Postprocessor report contains two DS8000 sections extent pool statistics and rank statistics These statistics are generated from SMF record 74 subtype 8 gt The extent pool statistics section provides capacity and performance information about allocated disk space For each extent pool it shows the real capacity and the number of real extents gt The rank statist
201. e a more dynamic response to changing workloads gt Provide simplified management of aliases gt Make it easier for users to make a decision to migrate to larger volume sizes In this section you see the commands and options that you can use for setup and control of HyperPAV and for the display of HyperPAV status information The migration to HyperPAV and the system requirements for HyperPAV are provided in this section 264 IBM System Storage DS8000 Host Attachment and Interoperability HyperPAV options The following SYS1 PARMLIB IECIOSxx options enable HyperPAV at the LPAR level HYPERPAV YES NO BASEONLY HyperPAV has the following options and corresponding actions YES Attempt to initialize LSSs in HyperPAV mode NO Do not initialize LSSs in HyperPAV mode BASEONLY Attempt to initialize LSSs in HyperPAV mode but only start I Os on base volumes The BASEONLY option returns the LSSs with enabled HyperPAV capability to a pre PAV behavior for this LPAR HyperPAV migration You can enable HyperPAV dynamically Because it can take time to initialize all needed LSSs in a DS8000 into HyperPAV mode planning is important If many LSSs are involved then pick a concurrent maintenance window with low I O activities to perform the SETIOS HYPERPAV YES command and do not schedule concurrent DS8000 microcode changes or IODF activation together with this change Example 12 7 shows a command example to dynamically activate HyperPAV Th
202. e ee eeee 123 7 1 vSphere introduction 0 0 0 0 c eee ee 124 7 2 vSphere storage concepts 00 e eee teens 125 TA NMES vss lad Sire Pei hea ee hase ST dd aed GP ae a a a a ai 125 72 2 Virtual CiSkSe oc sy ee ed ee ee a ee eee eg 135 7 2 3 Raw device mappings RDMs 0 eee 138 723 Multipathing 2 cc05 a6 d ee ee a ee ee ee Rie tt et eb eed eee 142 7 3 1 Pluggable Storage Architecture 0 0 0 0 ccc eens 142 43 2 Path WAMING a Ss ees Sethe Sead Geary ei oe eat leighton dene hae eta a anne 143 7 3 3 Multipathing and DS8000 0 eee 144 LA Storage VMotion ienesa naa na Bed wis e E a AENA soci DATARE 146 7 4 1 Steps to migrate a VM from one datastore to another 146 TA Eanos ioaea eE E E E Ee E EE EEN A 146 7 5 Using SSD volumes as Cache 0 teenies 147 TO Best PractiCeS aiisveo scons by A EA EEEE So Rota Sunk Go hele So beets ates 148 TOT DataStorest sini ond bag ha ban Poe eeb Bathe ee ee od Oeaube asad dees 148 76 2 MultipathinG aa ere EET cena ei pad ola eins ae bag tee eRe eas xia 4 ates 149 Chapter 8 Apple considerations 00 00 cece eee 151 8 1 Available resources 0 E T EER PEE ete eee 152 8 2 Configuring the Apple host on a DS8000 0 ce eee 152 8 3 Installing the ATTO software 0 0 c eee 152 8 3 1 ATTO OS X driver installation 000 0 ccc eee 152 8 3 2 ATTO Configuration Tool Installation
203. e gt x36501ab9 1s 1 dev mapper cut c 48 3600507630af fc29F0000000000004690 3600507630af fc29f0000000000004691 3600507630af fc29F0000000000004790 3600507630af fc29f0000000000004791 x36501ab9 partprobe x36501ab9 1s 1 dev mapper cut c 48 3600507630af fc29f0000000000004690 3600507630af fc29f0000000000004691 3600507 630af fc29f0000000000004691 part1 3600507630af fc29f0000000000004790 3600507630af fc29f0000000000004791 Example 6 32 was created with SLESS 11 The method works as well for Red Hat Enterprise Linux 5 but the partition names might be separate Older Linux versions One of the issues with older Linux versions is that LVM by default will not work with DM Multipath devices it does not exist in recent Linux versions 6 2 9 Attaching DS8000 Considerations This section provides additional considerations that specifically apply to the DS8000 Configuring DM multipath for DS8000 IBM provides a multipath conf file that contains the preferred settings for DS8000 attachment at the following website http www ibm com support docview wss rs 540 amp context S152G7 amp dc D430 amp ui d ssg1S4000 107 DM The relevant settings for DS8000 are shown in Example 6 33 Example 6 33 DM Multipath settings for DS8000 defaults polling_interval 30 failback immediate no_path_retry 5 path_checker tur user_friendly_names yes rr_min_io 100 devices 104 IBM System Storage DS8000 Host Attachment and
204. e information see the following website http msdn microsoft com en us library aa384649 28v vs 85 29 aspx Chapter 3 Windows considerations 33 3 9 3 Required components To use VSS functions you need an installed CIM client This CIM client requires the IBM System Storage DS command line interface CLI client to communicate with the DS8000 or ESSCLI for an ESS On each server the IBM API support is required for Microsoft Volume Shadow Copy Service as shown in Figure 3 12 Management server Figure 3 12 VSS installation infrastructure After the installation of these components which is described in IBM System Storage DS Open Application Programming Interface Reference GC35 0516 complete the following steps 1 Define a VSS_FREE volume group and virtual server 2 Define a VSS_RESERVED volume group and virtual server 3 Assign volumes to the VSS_FREE volume group The WWPN default for the VSS_FREE virtual server is 50000000000000 the WWPN default for the VSS_RESERVED virtual server is 50000000000001 These disks are available for the server as a pool of free available disks If you want to have separate pools of free disks you can define your own WWPN for another pool as shown in Example 3 12 34 IBM System Storage DS8000 Host Attachment and Interoperability Example 3 12 ESS Provider Configuration Tool Commands help C Program Files IBM ESS Hardware Provider for VSS gt ibmvssconfig exe
205. e information about protected and unprotected LUNs see 11 6 3 Protected versus unprotected volumes on page 221 gt Select the standard allocation method and rotate extents IBM System Storage DS8000 Host Attachment and Interoperability gt Select the volume sizes for IBM i For more information about IBM i volume sizes see 11 4 Sizing and numbering of LUNs on page 215 Figure 11 5 shows an example of creating an IBM i LUN Define Volume Characteristics Available capacity 1 544 GiB Projected remaining capacity 1 544 GiB Volume type Size GB 10 9 iSenes Protected l 141 12 a amp Volume quantity Storage allocation method Extent allocation method Standard m Rotate Extents v Optionally Choose Nickname Nickname prefix_ Nickname suffix Start ji Volume ID i Hexadecimal sequence Nickname example i01AB i01AC Optionally Assign Volume Groups To assign the new volume to a volume group select one or more volume groups and click OK To add another Create Volume Group action from the drop down mw iy eS h Select action w Ll Ba Select _ ali Sees Volumes a Figure 11 5 Defining an IBM i LUN If you use the crtfbvol DS CLI command to define IBM i LUNs specify the parameter 0S400 and the model of LUN that corresponds to a particular LUN size and protection Table 11 8 on page 215 For more information about protected and unprotected volumes see 11 6 3 Pr
206. e installation options When you have an unsupported driver version included with your Linux distribution either replace it immediately after installation or if it does not work at all use a driver disk during the installation This issue is currently exists for Brocade HBAs A driver disk image is available for download from the Brocade website see Loading the Linux Fibre Channel drivers on page 89 Important considerations gt Installing a Linux system on a SAN attached disk does not mean that it will be able to start from it You need to take additional steps to configure the boot loader or boot program gt Take special precautions when configuring multipathing if you are running Linux on SAN attached disks See 6 5 4 Booting Linux from DS8000 volumes on page 116 for details Initial Fibre Channel driver availability If the SAN attached DS8000 volumes are needed early in the Linux boot process for example if all or part of the system is located on these volumes it is necessary to include the HBA driver in the initial RAM filesystem initRAMFS image With the initRAMFS method the Linux book process provides certain system resources before the real system disk is set up The Linux distributions contain a script called mkinitrd that creates the initRAMFS image automatically Also the Linux distributions will automatically include the HBA driver if you already use a SAN attached disk during installation If not y
207. e modified and made unique to ensure data integrity Otherwise LVM can mix volumes from the source and the target sets A publicly available script called vgimportclone sh automates the modification of the metadata and thus supports the import of LVM volume groups that reside on a set of replicated disk devices 112 IBM System Storage DS8000 Host Attachment and Interoperability You can download this script from the CVS log for LVM2 scripts vgimportclone sh site at the following website http sources redhat com cgi bin cvsweb cgi LVM2 scripts vgimportclone sh cvsroot vm2 For an online copy of the script go to the UNIX Manual Page at the following website http www cl cam ac uk cgi bin manpage 8 vgimportclone Tip The vgimportclone script is part of the standard LVM tools for Red Hat Enterprise Linux 5 The SLES 11 distribution does not contain the script by default To ensure consistent data on the target volumes and avoid mixing up the source and target volume use a volume group containing a logical volume that is striped over two DS8000 volumes and FlashCopy to create a point in time copy of both volumes Then make both the original logical volume and the cloned volume available to the Linux system The DS8000 volume IDs of the source volumes are 4690 and 4790 the IDs of the target volumes are 46a0 and 47a0 The following steps demonstrate this procedure in detail 1 Mount the original file system using the LVM logi
208. e multipathing If you want to manage a multipathing design using VxVM including DMP you need to contact IBM and request a SCORE For more information about SCORE see 2 3 Additional supported configurations on page 13 10 6 1 HP UX multipathing solutions Up to HP UX 11iv2 PVLINKS was HP s multipathing solution on HP UX and was built into the LVM PVLINKS is a failover solution only which means that it performs a path failover to an alternate path if the primary path is not available PVLINKS does not offer load balancing which allows I O to be split among multiple paths while they are all available PVLINKS still exists in HP UX 1 1iv3 but IBM advises the use of native multipathing However information is provided in this section for instances where PVLINLKS is still used Multipathing Although HP PVLINKS is still available in HP UX 11iv3 IBM advises the use of native multipathing as a more complete method for implementing multiple paths to a DS8000 for both failover and load balancing To use PVLINKS for multipathing add the HP UX special device files that represent an additional path to the LUN on a newly created LVM volume group The first special device file for a disk device becomes the primary path and the other device files become alternate paths Before you can add the HP UX special device files to the LUN create a new LVM volume group the commands are illustrated in Example 10 7 1 Create a volume group a Crea
209. e paths as shown in Example 6 27 Example 6 27 Show multipath topology x36501ab9 multipathd k multipathd gt show topology 3600507630af fc29f0000000000004690 dm 0 IBM 2107900 size 50G features 1 queue_if_no_path hwhandler 0 wp rw policy round robin 0 prio 1 status active 3 0 0 0 sdb 8 16 active ready running Chapter 6 Linux considerations 101 3 0 1 0 sdf 8 80 active ready running 4 0 0 0 sdj 8 144 active ready running 4 0 1 0 sdn 8 208 active ready running 3600507630af fc29f0000000000004790 dm 1 IBM 2107900 size 50G features 1 queue_if_no_path hwhandler 0 wp rw policy round robin 0 prio 1 status active 3 0 0 2 sdd 8 48 active ready running 3 0 1 2 sdh 8 112 active ready running 4 0 0 2 sdl 8 176 active ready running 4 0 1 2 sdp 8 240 active ready running 3600507630af fc29f0000000000004791 dm 2 IBM 2107900 size 50G features 1 queue_if_no_path hwhandler 0 wp rw policy round robin 0 prio 1 status active 3 0 0 3 sde 8 64 active ready running 3 0 1 3 sdi 8 128 active ready running 4 0 0 3 sdm 8 192 active ready running 4 0 1 3 sdq 65 0 active ready running 3600507630af fc29f0000000000004691 dm 3 IBM 2107900 size 50G features 1 queue_if_no_path hwhandler 0 wp rw policy round robin 0 prio 1 status active 3 0 1 1 sdg 8 96 active ready running 3 0 0 1 sdc 8 32 active ready running 4 0 0 1 sdk 8 160 active ready running 4 0 1 1 sdo 8 22
210. e required for the VHD is allocated during creation This type is preferable for production workloads b Dynamically expanding The capacity for the VHD is allocated when the virtual machine writes to it This type provides a type of thin provisioning c Differencing This VHD is based on an existing virtual disk called parent The parent disk is read only and all changes to the differencing VHD are written into a separate file This type offers the possibility to revert changes What type of virtual hard disk do you want to create Fixed size This type of disk provides better performance and is recommended For servers running applications with high levels of disk activity The vhd file is created using the size of the fixed virtual hard disk It does not change when data is added or deleted Dynamically expanding This type of disk provides better use of physical storage space and is recommended for servers running applications that are not disk intensive The vhd file is small when the disk is created and grows as data is written to it C Differencing This type of disk is associated in a parent child relationship with another disk that you want to leave intact You can make changes to the data or operating system without affecting the parent disk so that you can revert the changes easily Figure 3 18 Virtual hard disk type 40 IBM System Storage DS8000 Host Attachment and Interoperability 5 Enter a filena
211. e that all 8 ports of the 2 node cluster are assigned to the same volume group V1 and therefore were assigned to the same set of LUNs Example 13 1 The Ishostconnect command output dscli gt lshostconnect Date Time 31 Mai 2012 14 04 33 CEST IBM DSCLI Version 7 6 20 221 DS IBM 2107 75TV181 Name ID WWPN HostType Profile portgrp volgrpID ITSO_SVC_N1_1 0000 500507680110455A SVC San Volume Controller 1 V1 ITSO_SVC_N1_2 0001 500507680120455A SVC San Volume Controller 1 V1 ITSO _SVC_N1_3 0002 500507680130455A SVC San Volume Controller 1 V1 ITSO_SVC_N1_4 0003 500507680140455A SVC San Volume Controller 1 V1 ITSO_SVC_N2_1 0004 5005076801104CD1 SVC San Volume Controller 1 V1 ITSO_SVC_N2_2 0005 5005076801204CD1 SVC San Volume Controller 1 V1 ITSO_SVC_N2_3 0006 5005076801304CD1 SVC San Volume Controller 1 V1 ITSO_SVC_N2_4 0007 5005076801404CD1 SVC San Volume Controller 1 V1 It is advisable to have one single volume group for the whole SVC cluster and to assign this volume group to all SVC ports Important Data corruption can occur if LUNs are assigned to both SVC nodes and non SVC nodes that is direct attached hosts The Managed Disk Link Count mdisk_link_count represents the total number of MDisks presented to the SVC cluster by that specific controller Chapter 13 IBM SAN Volume Controller considerations 283 13 3 3 DS8000 extent pool implications 284 In the DS8000 architecture extent pools are used to manage one or more
212. e the VTOC is large you can experience performance degradation An interface is implemented to return the highly allocated DSCB on volumes initialized with an indexed VTOC DFSMSadss uses this interface to limit VTOC searches and to improve performance The VTOC must be within the first 64 K 1 tracks but the index can be anywhere on the volume With LVs support it is possible to address a capacity of up to 3 964 PB where 64 K subchannels multiplied by 55 6 GB per volume equals to 3 64 PB To accommodate the needs of installations that require super large volumes IBM has developed an volume an extended address volume EAV called the 3390 Model A shown in Figure 12 2 Cylinders gt 65 520 Cylinders track managed lt or 65 520 space EAV Figure 12 2 Entire new EAV For the 3390 Model A support is enhanced to expand the volumes to 1 182 006 cylinders using 28 bit cylinder addressing which currently has a limit of 256 M tracks The existing CCHH track address became CCCCcccH x where the cylinder address is cccCCCC x and the head address is H x H 0 14 Starting with LMC level 7 6 30 xx the DS8000 supports up to Mod 1062 1 182 006 cylinders 1 004 TB Attention Copy Services are only allowed for 3390A with maximum size of Mod 236 x 3390 Model 1 262 668 cylinders Chapter 12 IBM System z considerations 255 The partial change from track to cylinder addressing creates two address areas on EAV vol
213. e time for I O with the same importance based on eWLM expectations for response time Tip I O Priority Manager allows multiple workloads on a single CKD volume with different priorities each other and permits the High important tasks always to have the best performance For more details on the use and configuration of I O Priority Manager see the Redbooks publication BM System Storage DS8000 Architecture and Implementation SG24 8886 and the Redpaper publication DS8000 I O Priority Manager REDP 4760 12 8 2 Easy Tier and I O Priority Manager coexistence Easy Tier is a workload balance activity on back end ranks and it works to balance workload on same ranks type and also to move high request tracks hot data to the more efficient ranks into the machine usually SSD For details see the Redbooks publication BM System Storage DS8000 Architecture and Implementation SG24 8886 I O Priority Manager works to balance I O activity on Host Adapter side limiting the access to the low priority hosts or tasks It means that while a rank is entering in saturation to avoid the possibility that all hosts and tasks that are using this rank will have a performance issue Important Easy Tier and I O Priority Manager can coexist because they work on different sides of the machine workload and both can improve machine performance 276 IBM System Storage DS8000 Host Attachment and Interoperability Figure 12 6 shows where Easy Tier and I O Pr
214. e volume size is 12288 MB equal to 12 GB To obtain the dmpnodename of this disk in VxVM use the vxdmpadm command used for Example 9 12 Example 9 12 VxVM commands before volume expansion bash 3 00 vxdmpadm getsubpaths ctlr c2 NAME STATE A PATH TYPE M DMPNODENAME ENCLR TYPE ENCLR NAME ATTRS NONAME DISABLED IBM_DS8x002_1 IBM_DS8x00 IBM_DS8x002 c2t50050763030CCO8Fd0s2 ENABLED A IBM_DS8x002_0 IBM_DS8x00 IBM_DS8x002 c2t50050763030CC1A5d0s2 ENABLED A IBM_DS8x001_0 IBM_DS8x00 IBM DS8x001_ c2t5005076303000663d0s2 ENABLED A IBM_DS8x000_0 IBM_DS8x00 IBM_DS8x000 bash 3 00 vxdisk list IBM_DS8x001_0 Device IBM_DS8x001_0 devicetag IBM DS8x001_0 type auto hostid v880 disk name IBM_DS8x001_0 id 1194446100 17 v880 group name 20781_dg id 1194447491 20 v880 info format cdsdisk privoffset 256 pubslice 2 privslice 2 flags online ready private autoconfig autoimport imported pubpaths block dev vx dmp IBM_DS8x001_0s2 char dev vx rdmp IBM_DS8x001_0s2 guid 9ecb6cb6 1dd1 11b2 af7a 0003ba43fdc1 udid 1BM 5F2107 5F7520781 5F6005076303FFC1A50000000000004704 site version 3 1 josize min 512 Bytes max 2048 blocks public slice 2 offset 65792 len 25095808 disk_offset 0 private slice 2 offset 256 len 65536 disk_offset 0 update time 1194447493 seqno 0 15 ssb actual_seqno 0 0 headers 0 240 configs count 1 len 48144 logs count 1 len 7296 Defined regions config priv 000048 000239 000192
215. ect the output into a file kextstat grep ATTO open terminal window to run command and redirect the output into a file java version open terminal window to run command and redirect the output into a file IBM System Storage DS8000 Host Attachment and Interoperability Solaris considerations This chapter provides information about the specifics of attaching IBM System Storage DS8000 systems to host systems running Oracle Solaris The following topics are covered Working with Oracle Solaris and DS8000 Locating the WWPNs of your HBAs Attaching Solaris to DS8000 Multipathing in Solaris Expanding dynamic volume with VxVM and DMP Booting from SAN vvvvvy Copyright IBM Corp 2011 2013 All rights reserved 159 9 1 Working with Oracle Solaris and DS8000 As with the previous models the IBM System Storage DS8000 continues to provide extensive support for the Solaris operating systems Currently the DS8000 supports Solaris 8 9 and 10 on a variety of platforms It also supports the VERITAS Cluster Server and the Oracle Solaris Cluster The IBM SSIC provides information about supported configurations including information about supported HBAs SAN switches and multipathing technologies The IBM SSIC website can be found at this website http www 03 ibm com systems support storage ssic interoperability wss Useful information about setting parameters of specific HBAs is also available in the IBM System Storage DS8
216. ed number of read commands simultaneously inflight It provides greater throughput over distance for IBM z OS Global Mirror XRC using the same distance Summary The Extended Distance FICON channel support produces performance results similar to XRC emulation mode at long distances It can be used to enable a wider range of choices in channel extension technology when using z OS Global Mirroring because emulation in the extender might not be required It can help reduce the total cost of ownership and provides comparable performance when mirroring application updates over long distances For more information see the Redbooks publication DS8000 Copy Services for IBM System z SG24 6787 268 IBM System Storage DS8000 Host Attachment and Interoperability 12 5 5 High Performance FICON for System z with multitrack support ZHPF To help increase system performance and provide greater System z FICON channel efficiency the DS8000 provides support for High Performance FICON for System z ZHPF with multitrack The ZHPF system with multitrack operations is an optional licensed feature on DS8000 storage systems ZHPF is supported starting from DS8000 licensed machine code 5 4 30 248 DS8000 bundle level 64 30 78 0 or higher With the introduction of ZHPF the FICON architecture was streamlined by removing significant impact from the storage subsystem and the microprocessor within the FICON channel A command block is created to chain commands i
217. edpaper publication IBM System Storage DS8700 Disk Encryption REDP 4500 280 IBM System Storage DS8000 Host Attachment and Interoperability 13 IBM SAN Volume Controller considerations This chapter describes the guidelines of how to attach an IBM System Storage DS8000 system to the IBM SAN Volume Controller SVC system The following topics are covered gt IBM System Storage SAN Volume controller gt SAN Volume controller multipathing gt Configuration guidelines Copyright IBM Corp 2011 2013 All rights reserved 281 13 1 IBM System Storage SAN Volume Controller The IBM System Storage SAN Volume Controller SVC is designed to increase the flexibility of your storage infrastructure by introducing an in band virtualization layer between the servers and the storage systems The SAN Volume Controller can enable a tiered storage environment to increase flexibility in storage management The SAN Volume Controller combines the capacity from multiple disk storage systems into a single storage pool It can be managed from a central point which is simpler to manage and helps increase disk capacity utilization It also allows you to apply SVC advanced Copy Services across storage systems from many vendors to help further simplify operations For more information about SAN Volume Controller see the Redbooks publication Implementing the IBM System Storage SAN Volume Controller V6 3 SG24 7933 13 2 SAN Volume Controller multipa
218. el A volumes are not supported at this time When running z OS as a virtual machine guest EAV volumes can be used if they are directly attached to the z OS guest The z VM operating system also provides the following support when using an FCP attachment gt FCP attached SCSI LUNs as emulated 9336 Model 20 DASD gt 1 TB SCSI LUNs 12 6 3 PAV and HyperPAV z VM support The z VM operating system provides PAV support This section provides basic support information for implementing PAV in a z VM environment You can find additional z VM technical information for PAV support at this website http www vm ibm com storman pav z VM guest support for dedicated PAVs The z VM operating system enables a guest z OS to use PAV and dynamic PAV tuning as dedicated volumes The following considerations apply gt Alias and base addresses must be attached to the z OS guest You need a separate ATTACH command for each alias address Attach the base address and its aliases address to the same guest gt A base address cannot be attached to a system if one of its alias addresses is attached to that guest It means that you cannot use PAVs for fullpack minidisks The QUERY PAV command is available for authorized class B users to query base and alias addresses with QUERY PAV rdev and QUERY PAV ALL Chapter 12 IBM System z considerations 273 gt To verify that PAV aliases are bound to the correct bases use the QUERY CHPID xx command combi
219. em Storage DS8000 Host Attachment and Interoperability Windows considerations This chapter provides details for attaching IBM System Storage DS8000 series systems to host systems for Windows Server 2003 and Windows Server 2008 The following topics are covered Attaching HBAs in Windows Installing SDD in Windows Clustering for Windows 2003 Server Using Multipath Input Output for Windows 2003 and 2008 Installing and configuring SDDDSM in Windows 2003 2008 Expanding dynamic disk for Windows 2003 Windows 2008 SAN boot support Windows Server 2003 Virtual Disk Service support Hyper V considerations vvvvvvvvy For additional information about the advantages disadvantages potential difficulties and troubleshooting with SAN booting see the Microsoft document Boot from SAN in Windows Server 2003 at the following addresses http technet microsoft com en us windowsserver bb512919 http technet microsoft com en us 1ibrary cc786214 28v WS 10 29 aspx Copyright IBM Corp 2011 2013 All rights reserved 17 3 1 Attaching HBAs in Windows DS8000 supports Fibre Channel attachment to Microsoft Windows Server 2003 and Windows Server 2008 servers For details regarding operating system versions and Host Bus Adapter HBA types see the IBM SSIC website http www 03 ibm com systems support storage ssic interoperability wss Attachment support includes cluster service The DS8000 can also act as a boot device Booting is supp
220. ent for the vxdiskadm utility is the VERITAS Enterprise Administrator VEA Figure 10 1 shows the presentation of disks using this GUI After creating the disk groups and the VxVM disks create file systems and mount them as shown in Figure 10 1 T L VERITAS Enterprise Administrator File Tools Actions Window E gt 4 EE B Connect Disconnect New Volu New Group Search fta System 8 Disks amp Diskview Volume Manager Alerts Management Console 2 4640 2 Disks A 4640 2 GA Cluster Nodes Device a Internal na Group na Poolname Status Size g Controllers S c2tod0 7 Not Initialized 0 gA Disk Groups External a yay Disks dg0101 dg01 gt Imported 11 997 GB gu Enclosures S c11t0d1 dg0102 dg01 Imported 11 997 GB File Systems S c12t0d1 Not Initialized 0 ql Saved Queries i a Volumes Control Panel Logs amp Network amp History 43 Favorite Hosts 4 W Normal usage High usage M Critical usage avr EH Console E 4640 2 m Figure 10 1 Disk presentation by VERITAS Enterprise Administrator 10 8 Working with LUNs This section provides information and examples regarding expanding LUNs as well as working with large LUNs The examples are based on a system that has a volume group named vg01 and a logical volume named 1vo11 10 8 1 Expanding LUNs First examine the characteristics of the volume group using the vgdisplay comm
221. er 2008 inside the parent partition The NTFS partitions can reside on any of these storage devices gt Local hard drives gt iSCSI LUNs gt Fibre Channel LUNs For information about how to create an NTFS partition on a Windows Server host see Chapter 3 Windows considerations on page 17 Hyper V offers two different ways to provide disk storage to virtual machines gt Virtual hard disks VHD VHDs are files that reside on an NTFS partition They offer extended features such as snapshots The maximum size of a VHD is 2 TiB gt Pass through disks Pass through disks are physical disks either local or SAN LUNs that are exclusively given to a virtual machine Access is routed through the parent partition Chapter 3 Windows considerations 37 The storage concept of Hyper V is outlined in Figure 3 14 Virtual machines can be equipped with either virtual IDE adapters or virtual SCSI adapters The differences are shown in Table 3 1 Table 3 1 virtual adapter types Adapter type Maximum Maximum number of Can a virtual machine number of disk drives per boot from this adapter adapters adapter virtual IDE 2 2 yes virtual SCSI 4 64 no Tip Hard disks can only be added to IDE adapters if the virtual machine is powered off If a special software package called integration services is installed inside the virtual machine both virtual IDE and virtual SCSI adapters provide the same performance The
222. er by using the text 5250 terminal mode interface with dedicated service tools DST or system service tools SST or with the System i Navigator GUI To add a logical volume in the DS8000 to the System ASP by using green screen SST complete the following steps 1 Run the STRSST command and sign on to start SST 2 Select option 3 Work with disk units Figure 11 7 System Service Tools SST Select one of the following Start a service tool Work with active service tools Work with disk units Work with diskette data recovery Work with system partitions Work with system capacity Work with system security Work with service tools user IDs ONAN AUNE Selection 3 F3 Exit F10 Command entry F12 Cancel Figure 11 7 System Service Tools menu Chapter 11 IBMiconsiderations 225 226 3 Select option 2 Work with disk configuration Figure 11 8 Work with Disk Units Select one of the following 1 Display disk configuration 2 Work with disk configuration 3 Work with disk unit recovery Selection 2 F3 Exit F12 Cancel Figure 11 8 Working with Disk Units menu When adding disk units to a configuration you can add them as empty units by selecting option 2 or you can select option 4 to allow IBM i to balance the data across all the disk units Balancing the data offers a more efficient use of the available resources 4 Select option 4 Add units to ASPs and balance dat
223. er considerations 57 4 3 Using NPIV with IBM VIOS and DS8000 58 VIOS 2 1 release with fix pack FP20 1 or later introduced virtual Fibre Channel capabilities NPIV to IBM Power Systems considerably extending the virtualization capabilities and enabling innovative solutions with IBM Power Systems and SAN storage NPIV is a standardized method for virtualizing a physical Fibre Channel port The PowerVM implementation of NPIV enables POWER LPAR to have virtual Fibre Channel adapters each with a dedicated WWPN With NPIV and unique WWPNs a system partition has its own identity in the SAN with direct access to both disk and tape SAN storage devices as shown in Figure 4 4 It is supported for example with the IBM 8 Gbps Fibre Channel HBA 5735 Tape Library RAA Fiber 8Gb PCle Chan areri ooooa Switch SAN Storage Figure 4 4 Virtual Fibre Channel support using NPIV on IBM Virtual I O Server IBM System Storage DS8000 supports VIOS with client partitions using virtual Fibre Channel adapters and NPIV This establishes techniques for SAN storage management such as SAN zoning LUN mapping and masking combined with NPIV to be used when provisioning dedicated storage to specific LPARs and applications without the need to additionally map those volumes as backing devices in the VIOS partition Furthermore solutions built around advanced storage system features such as copy services and bac
224. er installation macintosh 4 admin sudo kextstat grep ATTO 49 0 0x93a000 0x49000 0x48000 com ATTO driver ATl0CelerityFC8 1 4 4 lt 48 14 5 4 3 gt 8 3 2 ATTO Configuration Tool Installation and Flash Update Use the following procedure 1 Download the Mac OS X Config Tool file from the ATTO webpage as described under 8 3 1 ATTO OS X driver installation Unpack the file and perform a full installation For more information see the readme file shipped with the driver 3 The default path for installation is Applications ATTO Configuration Tool Launch the ATTO Configuration Tool double click the ATTO Configuration Tool located in the Applications ATTO Configuration Tool folder Check the installed Celerity Flash version to be sure it is the latest Expand host localhost select Celerity FC 8xEN Flash If there is a more current version available from the ATTO website download and unpack the new Celerity Flash contained in Mac OS X Flash Bundle Click Browse and select the current Celerity Flash bundle for your adapter card The Update button should become available Click Update to overwrite the adapter flash Aserver reboot is required for this update to take effect Set each channel NVRAM parameter Port Down Retry Count to 8 in case there are more than two ports or channels Set it to 30 if only one channel or port is available Do this by choosing host
225. ere There are multiple versions of VMware vSphere available However this chapter presents information based on vSphere Version 5 Most of this information also applies to earlier versions The following topics are covered vSphere introduction vSphere storage concepts Multipathing Storage vMotion SSD detection Best practices vvvvvy Copyright IBM Corp 2011 2013 All rights reserved 123 7 1 vSphere introduction VMware vSphere consists of several different components The major component is the ESXi server a bare metal hypervisor that runs on x86 servers ESXi provides a virtualization platform for a large variety of open systems such as Windows Linux or Solaris Before version 5 customers could choose between ESX and ESXi While ESX contains an integrated service console for management ESXi only consists of the hypervisor and requires the installation of external agents for monitoring and administration ESXi has a small footprint that allows it to be included in the server firmware Both ESX and ESXi provide the same capabilities in regard to virtual machines Because vSphere 5 and later will only contain ESXi this publication focuses on ESXi ESXi can connect to a variety of storage devices Local disks installed in the physical server Fibre Channel SAN iSCSI SAN NAS devices such as NFS shares vvvy For the management of vSphere VMware provides two components Figure 7 1 gt vCenter Server A manage
226. eries Within this book you can find information about the most popular host operating systems platforms including Windows IBM AIX VIOS Linux Solaris HP UX VMware Apple and IBM z OS The topics covered in this book target administrators or other technical personnel with a working knowledge of storage systems and a general understanding of open systems You can use this book as guidance when installing attaching and configuring System Storage DS8000 The practical usage oriented guidance provided in this book complements the IBM System Storage DS8000 Host Systems Attachment Guide SC26 7917 The team who wrote this book This book was produced by a team of specialists from around the world working for the International Technical Support Organization ITSO at the European Storage Competence Center ESCC in Mainz Axel Wesiphal is an IT Specialist for IBM Storage Systems at the IBM European Storage Competence Center ESCC in Mainz Germany He joined IBM in 1996 working for Global Services as a System Engineer His areas of expertise include setup and demonstration of IBM System Storage products and solutions in various environments Since 2004 Alex has been responsible for storage solutions and Proof of Concepts conducted at the ESSC with DS8000 SAN Volume Controller and IBM XIV He has been a contributing author to several XIV andDS8000 related IBM Redbooks publications Bertrand Dufrasne is an IBM Certified Consulting
227. ervers 9117 570 9119 590 and 9119 595 that allow IBM i and a partition Boot from SAN is supported on natively attached DS8000 and on DS8000 attached through VIOS NPIV or VIOS VSCSI With natively attached DS8000 the external load source requires IBM i level VSR3M5 or later Important Not all of the IBM i software levels are supported on a given hardware Table 11 1 on page 203 gives an overview of IBM i levels that can run on a particular hardware type For a complete and up to date list see the IBM SSIC website http www 03 ibm com systems support storage ssic interoperability wss The external load source unit can be connected with any of the IOAs listed in 11 2 Using Fibre Channel adapters on page 205 for native attachment When implementing boot from SAN with lOP based adapters attach to the boot from SAN IOP feature number 2847 to the connect the load source unit The IOAs that connect the other LUNs can be attached to IOPs feature number 2844 With native attachment to DS8000 multipath can be established for the load source LUN when connecting with IOP based or IOP less adapters on IBM i levels V6R1 or V7R1 However multipath is not possible for the load source LUN on IBM i V5R4 Boot from SAN is supported on any attachment to DS8000 with VIOS NPIV or VIOS VSCSI to IBM i When DS8000 is connected through VIOS to IBM i on POWER servers multipath will be established for the external load source which is connected t
228. es Policies uncheck Enable write caching on the Disk See Figure 3 1 m Computer Management Local om System Tools E Event Viewer Fa Shared Folders Local Users and Groups z Performance Logs and Alert a Sj Storage Device Manager Removable Storage gt Disk Defragmenter T Services and Applications F F Disk Management F volume Sic Partition Basic E amp E New Volume E Partition Basic C3New Volume F Partition Basic Disk 0 a Basic c 409 85 GB 409 85 GB NTF Online Healthy Systet IBM 2107900 Multi Path Disk Device Properties Disk 1 HE Sete Remove eraWere icomiithietaskianna Basic New ol or MB 86 MBN 950 ri R Online E mee z Healthy This setting improves disk performance but a po equipment failure might result in data loss or icp RoM 0 ggi i TY DYD D i File System Status NTFS Healthy System 409 85GB 405 37 GB NTFS Healthy 86 MB 84 MB NTFS Healthy 431MB 427MB General Policies Volumes Driver m Write caching and Safe Removal Gptimice for quick removal This setting disables wite cachi You Can disconnect this device icon ng Gn tHe disk andin ithout Using tHe sa pime for peronnence This setting enables write cachingiin windows tain Peformance lic disconnect this device tram te ca Figure 3 1 Shows how to disable the se
229. es not offer the feature of automatic load balancing 182 IBM System Storage DS8000 Host Attachment and Interoperability 10 1 3 Notes on naming conventions HP UX 11iv3 supports two different Volume Managers LVM The Logical Volume Manager LVM is the default Volume Manager for HP UX 11i More information about LVM can be found in the HP publication HP UX System Administrator s Guide Logical Volume Management VxVM The Veritas Volume Manager VxVM is a full featured Volume Manager and is included with HP UX 11i However there is only a base license provided that does not activate all of the features of this software More information about VxVM can be found in the Veritas Volume Manager Release Notes Important Each physical disk can only be managed by one Volume Manager at a time either LVM or VxVM but not both Both of these Volume Manager software packages can co exist on a HP UX 11i server each keeping track of the physical disks and logical volumes that it manages Both of these software packages are responsible for managing a pool of space and provisioning that space as logical containers According to the HP documentation these pools of space are referred to as volume groups by LVM VxVM refers to them as disk groups These two separate terms refer to the same concept the logical representation of one or more physical disks Tip LVM volume groups VxVM disk groups 10 1 4 Notes on enumeration Due to limitatio
230. es of direct I O and relieves the serialization of write accesses It helps improve performance for many environments particularly commercial relational databases In many cases the database performance achieved using CIO with JFS2 is comparable to that obtained by using raw logical volumes A method for enabling the concurrent I O mode is to use the mount o cio command when mounting a file system 5 6 4 Direct I O An alternative I O technique called direct I O bypasses the virtual memory manager VMM altogether and transfers data directly from the user s buffer to the disk and from the disk to the user s buffer The concept behind it is similar to raw I O in the sense that they both bypass caching at the file system level This reduces the CPU s processing impact and helps provide more memory available to the database instance which can make more efficient use of it for its own purposes Direct I O is provided as a file system option in JFS2 It can be used either by mounting the corresponding file system with the mount o dio command or by opening a file with the O_DIRECT flag specified in the open system call When a file system is mounted with the o dio option all files in the file system use direct I O by default Direct I O benefits applications that have their own caching algorithms by eliminating the impact of copying data twice first between the disk and the operating system buffer cache and then from the buffer c
231. ess gt Fibre Channel FCP attached SCSI devices The z VM system can also assign a Fibre Channel card running in FCP mode to a VM A Linux instance running in this VM can operate the card using the zfcp driver and access the attached DS8000 FB volumes To maximize the utilization of the FCP adapters share them between more than one VM However z VM cannot assign FCP attached volumes individually to virtual machines Each VM can theoretically access all volumes that are attached to the shared FCP adapter The Linux instances running in the VMs must ensure that each VM only uses the volumes that it is designated to use gt FCP attachment of SCSI devices through NPIV NPIV was introduced for System z z VM and Linux on System z to enable multiple virtual Fibre Channel HBAs running on a single physical HBA These virtual HBAs are assigned individually to virtual machines They log on to the SAN with their own WWPNs To the DS8000 they look exactly like physical HBAs You can create host connections for them and map volumes This method provides the ability to assign DS8000 volumes directly to the Linux virtual machine No other instance can access these even if it uses the same physical adapter card 6 2 2 Configuring Fibre Channel attachment In the following sections information about how Linux is configured to access DS8000 volumes is presented This section focuses on the Intel x86 platform but also presents information where ot
232. etas 172 9 6 1 Displaying the boot code 0 eee tee 173 9 6 2 Booting off a DS8000 LUN with Solaris 20 00 00 cece eee 174 9 6 3 Supplying the VID or PID string 0 0 0 c cee 175 9 6 4 Associating the MPxIO device file and underlying paths 179 Chapter 10 HP UX considerations 0 0 00 c cece ee 181 10 1 Working with HP UX c esius nania eia a teens 182 10 1 1 The agile View csini icare an eee SAA ees 182 10 1 2 Notes on multipathing a a auaa auaa aa 182 10 1 3 Notes on naming conventions 00 0 eee 183 10 1 4 Notes on enumeration 0 0 00 ce eae 183 10 2 Available resources 0 tees 183 10 3 Identifying available HBAS 00 0c tee 184 10 4 Identifying WWPNs of HBAS suasanana aaaea 184 10 5 Configuring the HP UX host for the DS8000 0000 eee 185 10 5 1 Device special files 2 0 c ec es 185 10 5 2 Device discovery aani EEEE a ees 185 10 6 Multipathiing rrena ae ie ines Sco ele oe E a Ow bial A elite latent 187 10 6 1 HP UX multipathing solutions 2 0 0 0 eee 187 10 6 2 Exposing link errors with HP UX 2 0 0 0 0 c ee eee 190 10 7 Working with VERITAS Volume Manager on HP UX 000 0c eee 191 10 8 Working with LUNS 0 0 2 teas 194 toS T Expanding LUNS ranno twat eae ates dette wd A EEEE AENEA 194 10 8 2 Working with large LUNS 0 0 eae 197 Chapter 11 IBM i considera
233. ete at ati doe kd bene ESEETO eet 244 11 9 1 Metro Mirror and Global Copy 0 0 00 cece eee 244 11 9 2 IBMidata migration 0 0 0 ce eae 244 Chapter 12 IBM System z considerations 00 00 cee eee eee 247 12 1 Connectivity considerations 0 0 c eee 248 TAA RICON kaaa Adon aaa a a E a eke Suh ow Werte a peels ane es 248 12 1 2 LINUX FCP connectivity 0 0 0 0 cee 248 12 2 Operating system prerequisites and enhancements 00 00e0e 248 12 3 Considerations for Z OS 0 0 ccc tees 249 12 3 1 Program enhancements for Z OS 0 0000s 249 12 3 2 DS8000 device definition saaa aaee 250 12 3 3 zDAC z OS FICON Discovery and Auto Configuration feature 251 12 3 4 Performance statistics 0 0 0 eee 253 12 3 5 Resource Measurement Facility 0 0 0c ee eee 253 12 3 6 System Management Facilities 0 00 00 eee 254 12 4 Extended Address Volume EAV Support 0 00 ee 254 12 4 1 Identifyingan EAV ane ee eae 257 12 4 2 z OS prerequisites for EAV 1 2 0 eee 259 12 4 3 EAV migration considerations aaa aaaea 260 12 5 FICON specifics for a z OS environment 0 000 cee eee 261 12 5 rl COVEIVIEW se hs ec Al ae toe ade fetes E Lalo et be ada Lt 261 12 5 2 Parallel access volumes PAV definition 0 0 00 262 12 5 3 HyperPAV z OS support and implementation 0 000000 264 12 5 4 Extended
234. evices such as queuing command reordering and so on Chapter 5 AIX considerations 77 5 6 2 Asynchronous I O Asynchronous I O operations run in the background and do not block user applications This improves performance because I O and application processing run simultaneously Many applications such as databases and file servers take advantage of the ability to overlap processing and I O although they need to take measures to ensure data consistency You can configure remove and change asynchronous I O for each device by using the chdev command or System Management Interface Tool SMIT Tip If the number of asynchronous I O requests is high then increase maxservers to approximately the number of simultaneous I Os that there might be In most cases it can be better to leave the minservers parameter at the default value because the asynchronous I O kernel extension will generate additional servers if needed When viewing the CPU utilization of the asynchronous I O servers if the utilization is even across all servers all servers are being used and you might want to try increasing their asynchronous I O number Running pstat a helps you view the asynchronous I O servers by name and running ps k shows them to you as the name kproc 5 6 3 Concurrent I O In 2003 IBM introduced a new file system feature called concurrent I O CIO for the second generation of the journaled file system JFS2 CIO includes all of the advantag
235. extends the reporting capability already available and provides a means of storing this history for extended periods of time for greater report capability Global Mirror enhancements TPC R V5 1 provides the ability to export Global Mirror history to a CSV file for RPO and Out of Sync Track analysis This feature extends the reporting capability already available and provides a means of storing this history for extended periods of time for greater report capability Enhanced session flexibility TPC R V5 1 for System z provides the ability to start a Global Mirror session in Global Copy mode This gives the user better control of replication services activity as DS8000 family storage systems begin forming consistency groups IBM WebSphere Application Services Tivoli Storage Productivity Center for Replication for System z supports IBM WebSphere Application Services for z OS V7 and V8 in 31 bit and 64 bit HyperSwap enhancements The new HyperSwap enhancements require APAR OA37632 Exploits features of the IBM DS8700 and DS8800 that allows HyperSwap to obtain information about the cause of storage replication suspension events and based on that information is able to allow systems to continue to run where it was not previously possible 12 9 2 References For more information about system requirements and operability see the following Redbooks publications gt IBM Tivoli Storage Productivity Center for Replication for System z V5 1 SC
236. f Resource Allocation Performance Co Status aaa ri Name SUSE VM E npiv Power Guest Snapshot E Open Console Bil Minrate Figure 7 25 Edit Settings of VM 2 Click Add Figure 7 26 I Show All Devices SUSE VM Virtual Machine Properties Hardware Options Resources Profiles vServices Gag e f Hardware Summary EE Memory 1024 MB m cPus 2 Video card Video card VMCI device Restricted E SCSI controller 0 LSI Logic Parallel Hard diski Virtual Disk CD DVD drive 1 Local_Disk _iso_imag ER Network adapter 1 VM Network amp Floppy drive Client Device Figure 7 26 Add device to VM Chapter 7 VMware vSphere considerations 135 3 From the list of devices in Figure 7 27 select Hard Disk Choose the type of device you wish to add Information This device can be added to this Virtual Machine QR SCSI Device Figure 7 27 Add Hard Disk 4 Select Create a new virtual disk as shown in Figure 7 28 A virtual disk is composed of one or more files on the host file system Together these files appear as a single hard disk to the quest operating system Select the type of disk to use Disk Create a new virtual disk Use an existing virtual disk Reuse a previously configured virtual disk c Figure 7 28 Create virtual disk Tip Select Use an existing virtual disk if you want to share v
237. f these methods you see all available disks In this example select Protect the data in this disk pool e Click OK New Disk Poot Type of disk pool Primary x Name of disk pool IpDSIASP cc tt lt S Database Benerated by the systernl Note Ifyou wantto create a switchable disk pool be sure to use the appropriate clustering function before using this wizard V Protect the data in this disk pool Cancel Help 2 Figure 11 17 Defining a new disk pool 230 IBM System Storage DS8000 Host Attachment and Interoperability 8 In the New Disk Pool Select Disk Pool window Figure 9 15 click Next New Disk Pool Select Disk Pool Log Selectthe existing disk pools to which you want to add disk units and click Add To remove a disk pool from the list select it and click Remove Click New Disk Pool to create a new disk pool Delete new disk pools by selecting the disk pool and clicking Delete Existing disk pools Add disk units to these disk pools Disk Pool _ Type Of F Add gt Disk Pool Type Of Protection Basic ee Ds_ia Primary Protected New Disk Pool wj Figure 11 18 New Disk Pool Select Disk Pool panel 9 In the Disk Pool New Disk Pool Add Disks Units window Figure 11 19 click Add Parity Protected Disks The disks or LUNs to be added can be partially protected Disk Pool New Disk Pool Add Disks Units Disk pool Ds_ias
238. fit from ZHPF multitrack function In addition ZHPF complements the EAV for System z strategy for growth by increasing the I O rate capability as the volume sizes expand vertically In laboratory measurements multitrack operations for example reading 16 x 4 KB per I O converted to the ZHPF protocol on a FICON Express8 channel achieved a maximum of up to 40 more MBps than multitrack operations using the native FICON protocol IBM laboratory testing and measurements are available at the following website http www ibm com systems z hardware connectivity FICON performance html The ZHPF with multitrack operations is available on z OS V1 9 or higher with the PTFs for APARs OA26084 and OA29017 Currently ZHPF and support for multitrack operations is available for the IBM System z10 z196 and z114 servers and applies to all FICON Express8S available only on z196 and z114 FICON Express8 FICON Express4 and FICON Express2 features CHPID type FC The FICON Express adapters are not supported Chapter 12 IBM System z considerations 269 The ZHPF is transparent to applications however z OS configuration changes are required for example HCD must have CHPID type FC defined for all the CHPIDs that are defined to the DS8000 control unit For the DS8000 installation of the licensed feature key for the High Performance FICON feature is required When these items are addressed existing FICON port definitions in the DS8000 will funct
239. for the IBM System Storage DS8000 series system attachment to IBM i The following topics are covered Supported environment Using Fibre Channel adapters Sizing and numbering of LUNs Using multipath Configuration guidelines Booting from SAN Installing IBM i with boot from SAN through VIOS NPIV Migrating vvvvvvvy For more information about these topics see BM i and IBM System Storage A Guide to Implementing External Disks on IBM i SG24 7120 Copyright IBM Corp 2011 2013 All rights reserved 201 11 1 Supported environment This section describes the hardware and software prerequisites for attaching the DS8000 to an IBM i system Scope This chapter provides a high level summary of needed servers and software levels to connect DS8000 to the IBM i For detailed hardware and software prerequisites see the System Storage Interoperation Center SSIC and the BladeCenter Interoperability Guide BIG listed in 11 1 4 Useful websites on page 204 The following attachment methods are supported for IBM i gt Native attachment by connecting to the DS8000 using physical adapters in IBM i note that IBM i resides in a partition of a POWER system or in a former System i model gt Attachment with virtual input and output server VIOS node port ID virtualization NPIV gt Attachment with VIOS using virtual SCSI adapters 11 1 1 System hardware Native connection is supported on IBM POWER7 IBM POWER6 and POW
240. g Easy Tier the IBM i based options for hot spot management for an IBM i workload are described in the following section Chapter 11 IBM i considerations 211 Using IBM i tools and methods Alternatively to Easy Tier and STAT tool you can use the IBM i tools for hot spot management IBM i data relocation methods can be used with natively and VIOS_NPIV connected DS8000 because IBM i recognizes the LUNs that reside on SSD in DS8000 by the LUN Vital Product Data VPD The following IBM i tools and methods are available to you gt Performance Explorer PEX is a data collection tool in IBM i which collects information about a specific system process or resource to provide detailed insight PEX can be used to identify the IBM i objects that are the most suitable to be relocated to solid state drives For this usually a PEX collection of IBM i disk events is run such as synchronous and asynchronous reads synchronous and asynchronous writes page faults and page outs gt iDoctor is a suite of tools used to manage the collection investigate and analyze performance data on IBM i The data collected by PEX can be analyzed by the PEX Analyzer tool of iDoctor An example of iDoctor showing read operations by IBM i objects can be seen in Figure 11 3 You might notice that big majority of read operations goes to the three IBM i objects so these objects are good candidates to relocate to SSD I O reads by object 10000 L co 8000
241. g of open systems You can use this book as guidance when installing attaching and configuring System Storage DS8000 The practical usage oriented guidance provided in this book complements the BM System Storage DS8000 Host Systems Attachment Guide SC26 7917 SG24 8887 01 ISBN 073843759X ll l Ihi ul qu A We Redbooks INTERNATIONAL TECHNICAL SUPPORT ORGANIZATION BUILDING TECHNICAL INFORMATION BASED ON PRACTICAL EXPERIENCE IBM Redbooks are developed by the IBM International Technical Support Organization Experts from IBM Customers and Partners from around the world create timely technical information based on realistic scenarios Specific recommendations are provided to help you implement IT solutions more effectively in your environment For more information ibm com redbooks
242. ge datastore Move the virtual machine s storage to another datastore Figure 7 46 Change datastore 4 Select the target datastore from the list and click Next 5 In the Ready to Complete window click Finish The duration of the migration depends on several factors such as workload size of VM number of available paths storage device and so on 7 4 2 Limitations Storage vMotions require a host to read all data from the VM on one datastore and write the data to another This generates additional load on the storage adapters and paths which need to be taken into consideration when performing Storage vMotions vSphere limits the amount of parallel Storage vMotion operations as listed in Table 7 1 146 IBM System Storage DS8000 Host Attachment and Interoperability Table 7 1 Storage vMotion limits Operation Limit Concurrent Storage vMotion operations per datastore 8 Concurrent Storage vMotion operations per host 2 7 5 Using SSD volumes as cache ESXi allows for overcommitment of resources including memory In times of high memory usage by the VMs the ESXi host could run out of physical memory and the hypervisor needs to swap out memory pages These memory pages are stored on a swapfile residing on physical disks or SAN volumes ESXi can detect volumes that reside on SSD ranks automatically Such volumes can be used as cache by the ESXi host Figure 7 47 shows how such volumes are identified in the S
243. ght bytes DS8000 LUNs for the host type zLinux are created from the DS8000 volume ID as demonstrated in Example 6 19 Example 6 19 DS8000 LUN format for zLinux dscli gt showvolgrp lunmap V18 Name ITSO_zLinux ID V18 Type SCSI Mask Vols 4800 4900 4901 4902 4800 40484000 4900 40494000 4901 40494001 4902 40494002 Using the Isscsi command The Isscsi command produces odd LUNs for DS8000 volumes in Linux on System z These LUNs are decimal representations of word swapped DS8000 LUNs For example the 1sscsi command shows LUN 1073758281 which is 0x40004049 hexadecimal and corresponds to DS8000 LUN 0x40494000 Chapter 6 Linux considerations 97 Red Hat Enterprise Linux uses the etc zfcp conf file to configure SAN attached volumes The format of this file is shown in Example 6 20 The three bottom lines in this example are comments that explain the format Comments are not required in the file Example 6 20 Format of the etc zfcp conf file for Red Hat Enterprise Linux Inxmnt01 cat etc zfcp conf 0x0500 0x500507630a00029f 0x4049400000000000 0x0500 0x500507630a00029f 0x4049400100000000 0x0500 0x500507630a0b029f 0x4049400000000000 0x0500 0x500507630a0b029f 0x4049400100000000 0x0600 0x500507630a13029f 0x4049400000000000 0x0600 0x500507630a13029f 0x4049400100000000 0x0600 0x500507630a1b029f 0x4049400000000000 0x0600 0x500507630a1b029f 0x4049400100000000 FCP HBA LUN Remote DS8000 Port 6 2 8 Setting up devi
244. gured in a failover cluster gt All hosts involved have access to the LUN that the shared NTFS resides on gt The same drive for the volume letter is used on all hosts Access to the CSV is controlled by a coordinator node that is chosen automatically Every node can directly write to the volume by its iSCSI SAS or Fibre Channel connection but file system metadata is routed over a private network to the coordinator node that makes the required file system metadata updates If the coordinator node fails the role is moved to any of the remaining nodes For more information about CSV see this website http technet microsoft com en us 1ibrary dd630633 28WS 10 29 aspx 44 IBM System Storage DS8000 Host Attachment and Interoperability 3 10 6 Best practices Device management of Hyper V is done by the Windows Server 2008 Therefore all steps required to attach DS8000 LUNs to a Hyper V server are identical to those required for Windows Server 2008 It includes installing the correct HBA drivers and a multipathing driver such as SDDDSM To avoid virtual machines impacting the performance of the overall Hyper V installation it is advisable to store VHD files on dedicated NTFS partitions on separate LUNs VHD files can be migrated between partitions using the regular Windows copy and move functions The VHD file however must not be in use during this operation Chapter 3 Windows considerations 45 46 IBM System Storage DS8000 Host A
245. hable or private entities To add LUNs or create private non switchable IASPs using the Operations Navigator GUI complete the following steps 1 Start System i Navigator Figure 11 13 shows the initial window System i Navigator File Edit View Help My Connections Signe Description fi 9 155 57 42 Qsecofr v7rimO ITSO IBMi fi 9 155 70 14 v rimO Mainz I6PFE Environment My Connections Management Central 9 155 70 40 fa My Connections 9 155 57 42 Release i pea fil 9 155 70 15 v5r4m0 AS400PFE fi pee tamas ee a z f pp fil 9 155 70 40 v7r im Manage this system fi E A il MceiSl1 main v5r4m0 Manage this system Add a connection ii Change all of your system passwords Install plug ins Figure 11 13 System i Navigator initial window 228 IBM System Storage DS8000 Host Attachment and Interoperability 2 Click the expand icon to the server that you want to add the logical volume to and sign on to that server as shown in Figure 11 14 System i signon iil Signonto 89 155 57 42 User ID QSECOFR Password Cancel Figure 11 14 System i Navigator Signon to System i window 3 As shown in Figure 11 15 expand Configuration and Service gt Hardware Disk Units Environment My Connections My Connections Name Release Description 92 155 57 42 vimo ITSOi 92 155 70 14 v rim0 Mainz I6PFE Bf 9 155 57 42 a
246. hannel 20 00 00 c0 dd 18 3b 7b 21 00 00 c0 dd 18 3b 7b ISP2532 based 8Gb Fibre Channelto PCI Express HBA vmhba2 Fibre Channel 20 00 00 24 ff 2d c3 40 21 00 00 24 ff 2d c3 40 Figure 7 6 Rescan devices 7 Verify that all volumes are correctly detected by selecting each vmhba connected to the DS8000 The detected volumes are listed in the details see Figure 7 7 Details vmhba2 Model ISP2532 based 8Gb Fibre Channel to PCI Express HBA WWN 20 00 00 24 ff 2d c3 40 21 00 00 24 ff 2d c3 40 Targets 2 Devices 2 Paths 4 View Devices Paths Name Runtime Name LUN Type Drive Type Transport Capacity IBM Fibre Channel Disk naa 600507630affc29f wvmhba2 C0 TO LO 0 disk Non SSD Fibre Chann 5 00 GB IBM Fibre Channel Disk naa 600507630affc29f wmhba2 C0 TO L1 1 disk Non SSD Fibre Chann 20 00 GB Figure 7 7 detected volumes 8 To create a new datastore on the detected volumes go to Configuration Storage and click Add storage as shown in Figure 7 8 This action starts a wizard that guides you through the steps to create a new VMFS datastore View Datastores Devices Datastores Refresh Delete Add Storage Rescan All Identification Status Drive Type Capacity Free Type amp Local_Disk Normal LSILOGIC Serial A Non SSD 196 25 GB 64 69 GB VMFSS Device Figure 7 8 Add Storage a In Figure 7 9 select storage type Disk LUN and click Next
247. hannels limit the number of devices per channel to 16 384 To fully access 65 280 devices on a storage unit it is necessary to connect a minimum of four FICON host channels to the storage unit This way by using a switched configuration you can expose 64 control unit images 16 384 devices to each host channel Chapter 12 IBM System z considerations 261 Quick Init for z OS Starting from release 6 2 LMC level 7 6 20 xx volume initialization is performed dynamically after volume creation or expansion There is no delay for initialization before going online to the host The customer can initiate copy services operations immediately after volume creation or expansion without needing to wait for full volume initialization It addresses special requirements of accounts that re provision their volumes regularly Multiple Allegiance Normally if any System z host image server or LPAR does an I O request to a device address for which the storage disk subsystem is already processing an I O that came from another System z host image the storage disk subsystem will send back a device busy indication This delays the new request and adds to the overall response time of the I O this delay is shown in the Device Busy Delay AVG DB DLY column in the RMF DASD Activity Report Device Busy Delay is part of the Pend time The DS8000 series accepts multiple I O requests from different hosts to the same device address increasing parallelism and reducing
248. hat each path to logical volumes uses a separate VIOS separate SAN switches and host adapters in separate I O drawer pairs in the DS8000 With such implementation IBM i will keep resiliency even if a VIOS fails 11 5 2 Configuring the multipath 218 To achieve multipathing when DS8000 attaches natively to IBM i assign the volumes to two or more ports from Fibre Channel adapters or IBM i host adapters When connecting with VIOS NPIV attach DS8000 to an IBM i LPAR with two VIOS in NPIV and assign the LUNs to the virtual Fibre Channel adapter that connect to both of the VIO servers to provide availability in case of maintenance or failure of one VIO servers You can also connect IBM i with more than two VIOS and setup one path to the LUNs through each of them to provide even better resiliency Up to eight paths can be established to a LUN for IBM i In certain installations each VIOS connects to a separate SAN such as DS8000 attaching to both SANs In this case consider connecting two virtual Fibre Channel adapters to each VIOS Each virtual Fibre Channel adapter is assigned to a separate physical adapter and communicates to a separate SAN This way you help ensure resiliency in SANs during outage of one VIOS In the DS8000 assign the LUNs to all four virtual Fibre Channel adapters in IBM i IBM System Storage DS8000 Host Attachment and Interoperability For example a client is performing maintenance on one VIOS If the other ac
249. hd k show top map 3600507630affc29f0000000000004791 3600507630af fc29f0000000000004791 dm 4 IBM 2107900 size 100G features 1 queue_if_no_path hwhandler 0 wp rw policy round robin 0 prio 1 status active 3 0 0 3 sde 8 64 active ready running 1 3 sdi 8 128 active ready running 3 sdq 65 0 active ready running 3 0 4 0 1 4 0 0 3 sdm 8 192 active ready running Resize the file system and check the new capacity as shown in Example 6 50 Example 6 50 Resizing the file system and check capacity x36501ab9 resize2fs dev mapper 3600507630affc29f0000000000004791 resize2fs 1 41 9 22 Aug 2009 Filesystem at dev mapper 3600507630af fc29f0000000000004791 is mounted on mnt itso_4791 on line resizing required old desc_blocks 4 new_desc_ blocks 7 Performing an on line resize of dev mapper 3600507630affc29f0000000000004791 to 26214400 4k blocks IBM System Storage DS8000 Host Attachment and Interoperability The filesystem on dev mapper 3600507630af fc29f0000000000004791 is now 26214400 blocks long x36501ab9 df h mnt itso_4791 Filesystem Size Used Avail Use Mounted on dev mapper 3600507630af fc29f0000000000004791 99G 7 9G 86G 9 mnt itso_4791 Dynamic volume increase Currently the dynamic volume increase process has the following restrictions gt From the supported Linux distributions only SLES11 SP1 and RHEL 6 have this capability gt The sequence works only with unpartitio
250. he 1sscsi output is the internal Linux SCSI address It is for historical reasons constructed like a traditional parallel SCSI address It consists of four fields gt HBA ID Each HBA in the system whether parallel SCSI FC or even a SCSI emulator gets a host adapter instance when it is initiated gt Channel ID This ID is always zero It was formerly used as an identifier for the channel in multiplexed parallel SCSI HBAs gt Target ID For parallel SCSI it is the real target ID the one set through a jumper on the disk drive For Fibre Channel it represents a remote port that is connected to the HBA With it you can distinguish between multiple paths and between multiple storage systems gt LUN This item is rarely used in parallel SCSI In Fibre Channel it is used to represent a single volume that a storage system offers to the host The LUN is assigned by the storage system Figure 6 1 illustrates how the SCSI addresses are generated Usually 0 for parallel scsl Tt Ff T T HBA ID Channel ID Target ID LUN Figure 6 1 Understanding the composition of Linux internal SCSI addresses IBM System Storage DS8000 Host Attachment and Interoperability 6 2 6 Identifying DS8000 devices The udev subsystem creates device nodes for all attached devices In the case of disk drives it not only sets up the traditional dev sdx nodes but also other representatives such as the ones that can be found in dev di
251. he CNAs is called bfa You can download the driver from this website http www brocade com services support drivers downloads adapters index page gt IBM FICON Express The HBAs for the System z platform They can either operate in FICON for traditional CKD devices or FCP mode for FB devices Linux deals with them directly only in FCP mode The driver is part of the enterprise Linux distributions for System z and is called zfcp Kernel modules or drivers are loaded with the modprobe command They can also be removed as long as they are not in use as shown in Example 6 3 Example 6 3 Load and unload a Linux Fibre Channel HBA kernel module x36501ab9 modprobe qla2xxx x36501ab9 modprobe r qla2xxx Upon loading the FC HBA driver examines the FC fabric detects attached volumes and registers them in the operating system To see whether a driver is loaded and the dependencies for it use the 1smod command Example 6 4 Example 6 4 Filter list of running modules for a specific name x36501ab9 1smod tee gt head n 1 gt grep qla gt dev null Module Size Used by qla2xxx 293455 0 scsi_transport_fc 54752 1 qla2xxx scsi_mod 183796 10 qla2xxx scsi_transport_fc scsi_tgt st ses With the modinfo command you see detailed information about the kernel module itself such as the version number what options it supports and so on You can see a partial output in Example 6 5 Example 6 5 Detailed information abou
252. he DS CLI on this host and execute the Ishostvol command Example 9 10 shows the output Example 9 10 Ishostvol output bash 3 00 opt ibm dscli bin lshostvol sh Device Name Volume ID c2t50050763030CC1A5d0 c2t50050763030CC08Fd0 c2t5005076303000663d0 c3t50050763030B0663d0 IBM 2107 7520781 4704 IBM 2107 7503461 4704 IBM 2107 75ABTV1 4704 IBM 2107 75ABTV1 4704 c3t500507630319C08Fd0 IBM 2107 7503461 4704 c3t50050763031CC1A5d0 IBM 2107 7520781 4704 In Example 9 10 the volume with ID 75207814704 is c2t50050763030CC1A5d0 or c3t50050763031CC1A5d0 on the Solaris host To see the size of the volume on the Solaris host use the luxadm command as shown in Example 9 11 Example 9 11 luxadm output before volume expansion bash 3 00 luxadm display dev rdsk c2t50050763030CC1A5d0s2 DEVICE PROPERTIES for disk dev rdsk c2t50050763030CC1A5d0s2 Status Port A 0 K Vendor IBM Product ID 2107900 WWN Node 5005076303ffcla5 WWN Port A 50050763030ccla5 Revision 991 Serial Num 75207814704 Unformatted capacity 12288 000 MBytes Write Cache Enabled Read Cache Enabled Minimum prefetch 0x0 Maximum prefetch 0x16 Device Type Disk device Path s Chapter 9 Solaris considerations 167 dev rdsk c2t50050763030CC1A5d0s2 devices pci 9 600000 SUNW q1c 1 fp 0 0 ssd w50050763030cc1a5 0 c raw dev rdsk c3t50050763031CC1A5d0s2 devices pci 9 600000 SUNW q1c 2 fp 0 0 ssd w50050763031cc1a5 0 c raw This indicates that th
253. he RAID ranks and device adapters Basically I O Priority Manager uses QoS to assign priorities for different volumes and applies network QoS principles to storage by using a particular algorithm called Token Bucket Throttling for traffic control I O Priority Manager is designed to understand the load on the system and modify it by using dynamic workload control 12 8 1 Performance groups For System z there are 14 performance groups three performance groups for high performance policies four performance groups for medium performance policies six performance groups for low performance policies and one performance group for the default performance policy Only with System z two operation modes are available for I O Priority Manager without software support or with software support In the Workload Manager WLM Service Definition the service option I O priority management must be set to Yes to being able to use it See also Figure 12 6 on page 277 to verify where it can be enabled on DS8000 Support is provided for CKD I O priority with software input gt The user assigns a performance policy to each CKD volume that applies in the absence of additional software support gt The OS can optionally specify parameters that determine priority of each I O operation gt In general the priorities will be managed by z OS through eWLM gt Itis supported on z OS V1 11 V1 12 V1 13 and above I O Priority Manager is a DS8000
254. he Storage Tier Advisory Tool To size the number of Solid State Drives for an IBM i workload consider using Storage Tier Advisor Tool STAT for IBM i LUNs STAT provides you the heat distribution of on the IBM i data Figure 11 1 shows an example of heat distribution on IBM i Volume Heat Distribution Capacity on Tier 7 Heat Distribution 8 bec 65G 66G 64G 666 65G 66G 65G 66G 65G 66G 65G 66G 66G 65G 66G 65G 66G 64G le ge 2 gt al Displaying Page 3 of 7 Figure 11 1 STAT output for an IBM i workload For more information how to use STAT see the Redpaper publication BM System Storage DS8000 Easy Tier REDP 4667 02 Using guidelines For sizing SSDs at the installations who do not have Easy Tier installed and therefore cannot use STAT you might want to use the following guidelines gt Consider a rough guideline that one SSD replaces 5 15 K RPM HDD as far as performance of IBM i workload are concerned IBM System Storage DS8000 Host Attachment and Interoperability gt Consider the skew level shown in Figure 11 2 it presents the percentage of workload small size I O on the percentage of active data It is a skew level of an IBM i benchmark workload that is a good emulation of a general IBM i customer s workload 100 90 80 4 70 60 50 40 Percentage of workload 0 10 20 30 40 50 60 70 80 90 100 Percentage of activedata Percent of small O
255. her platforms differ Most examples shown are command line based Loading the Linux Fibre Channel drivers The following are four main brands of Fibre Channel Host Bus Adapters FC HBAs gt QLogic These adapters are the most used HBAs for Linux on the Intel X86 platform There is a unified driver for all types of QLogic FC HBAs The name of the kernel module is gla2xxx and is included in the enterprise Linux distributions The shipped version is supported for DS8000 attachment gt Emulex These adapters are sometimes used in Intel x86 servers and rebranded by IBM the standard HBA for the Power Systems platform There also is a unified driver that works with all Emulex FC HBAs The kernel module name is pfc A supported version is also included in the enterprise Linux distributions for both Intel x86 and Power Systems gt Brocade Converged network adapters CNA that operate as FC and Ethernet adapters and are relatively new to the market They are supported on the Intel x86 platform for FC attachment to the DS8000 The kernel module version provided with the current enterprise Linux distributions is not supported You must download the supported version from the Brocade website The driver package comes with an installation script that compiles and installs the module Chapter 6 Linux considerations 89 There might be support issues with the Linux distributor because of the modifications done to the kernel The FC kernel module for t
256. hod similar to the attachment Write the LUN of the volume into the unit_remove metafile for each remote port in sysfs Tip For newly added devices to be persistent use the methods provided in 6 2 7 Adding DS8000 volumes to Linux on System z on page 96 to create the configuration files to be used at the next system start 6 2 11 Adding new DS8000 host ports to Linux on System z If you connect new DS8000 ports or a new DS8000 system to Linux on System z complete the following steps to logically attach the new remote ports 1 Discover DS8000 ports that are connected to the HBAs as shown in Example 6 42 108 IBM System Storage DS8000 Host Attachment and Interoperability Example 6 42 Discovering connected remote ports Inxmnt01 zfcp_san_disc Ww b 0 0 0500 0x500507630a00029F 0x500507630a0b029F Inxmnt01 zfcp_san_disc Ww b 0 0 0600 0x500507630a03029f 0x500507630a08029f 2 Attach the DS8000 ports logically to the HBAs As Example 6 43 shows there already is a remote port attached to HBA 0 0 0500 It is the one path already available to access the DS8000 volume Example 6 43 Listing and attaching remote ports Inxmnt01 1s sys bus ccw devices 0 0 0500 grep Ox 0x500507630a00029f Inxmnt01 echo 0x500507630a0b029f gt sys bus ccw devices 0 0 0500 port_add Inxmnt01 1s sys bus ccw devices 0 0 0500 grep Ox 0x500507630a00029f 0x500507630a0b029f 3 Add the second connected DS8000 port to the HBA
257. ht have more than one individual with access to any single piece of information that is required to determine an encryption key When this method is used you can configure redundant key servers Redundant key servers are two or more key servers that have independent communication paths to encrypting storage devices You must create at least two key servers before you can create an encryption group For more information about redundancy and deadlock prevention see the Encryption concepts section of the Information Center Starting from release 6 2 LMC level 7 6 20 xx on z OS environments the security key can be also managed internally by z OS hosts using ISKLM 1 1 It will provide same functionality provided by TKLM servers as described previously without needing machines external to the host Nothing has changed from the DS8000 point of view The new version of key server software for zSeries platform is based off the original encryption key manager EKM that was used for tapes gt ISKLM 1 1 provides a GUI interface similar to TKLM gt It provides a migration path for zSeries EKM Customers to manage security key without needing to have TKLM servers Encryption The DS8800 now has a full range of encrypted disks including SSD 400 GB that permits to have Easy Tier fully operational also on encrypted machines For more detailed information see the Redbooks publication BM System Storage Data Encryption SG24 7797 and the R
258. ice SDDDSM command as shown in Example 3 8 Example 3 8 The datapath query device command before expansion C Program Files IBM SDDDSM gt datapath query device Total Devices 3 DEV 0 DEVICE NAME Disk2 PartO TYPE 2107900 POLICY OPTIMIZED SERIAL 75034614703 Path Adapter Hard Disk State Mode Select Errors 0 Scsi Port1 BusO Disk2 Part0 OPEN NORMAL 42 0 1 Scsi Port2 Bus0 Disk2 Part0 OPEN NORMAL 40 0 DEV 1 DEVICE NAME Disk3 PartO TYPE 2107900 POLICY OPTIMIZED SERIAL 75207814703 Path Adapter Hard Disk State Mode Select Errors 0 Scsi Port1 Bus0 Disk3 Part0 OPEN NORMAL 259 0 1 Scsi Port2 Bus0 Disk3 Part0 OPEN NORMAL 243 0 DEV 2 DEVICE NAME Disk4 PartO TYPE 2107900 POLICY OPTIMIZED SERIAL 75ABTV14703 Path Adapter Hard Disk State Mode Select Errors 0 Scsi Port1 Bus0 Disk4 Part0 OPEN NORMAL 48 0 1 Scsi Port2 Bus0 Disk4 Part0 OPEN NORMAL 34 0 In Example 3 8 you can see that the volume with ID 75207814703 is Disk3 on the Windows host because the volume ID matches the SERIAL You can see the size of the volume by using Disk Management Figure 3 7 Computer Management ioj x File Action View Window Help li x e Amem peg Volume Layout _ Type File System Status Capacity Free Space Free Computer Management Local j K System Tools J c Partition Basic NTFS Healthy Boot 68 36 GB 62 00 GB 90 o fal Event Viewer amp
259. ich are enhance with SDDDSM because it has module integration and works closer to the OS Windows Server 2008 builds upon that and works natively by integrating MPIO and further supporting IBM disk drive configurations Make sure to read the differences in the following section links for SDD and SDDDSM 3 2 Installing SDD in Windows An important task with a Windows host is the installation of the SDD multipath driver Ensure that SDD is installed before adding additional paths to a device otherwise the operating system can lose the ability to access existing data on that device For details see the BM System Storage Multipath Subsystem Device Driver User s Guide SC30 4131 To make sure the SDD system is installed you can verify if its installation files exist under C Program Files ibm SDD SDD and all the required guide documentation can be downloaded from the following link http www ibm com support docview wss uid ssg1S4000054 18 IBM System Storage DS8000 Host Attachment and Interoperability Support Windows 2000 is no longer supported This chapter includes SDD information about Windows Server 2003 and 2008 Tip During the SDD installation the option Cache on Disk might be enabled in which case the system will use the server s cache The DS8k already has cache enabled therefore it is not necessary to use the servers cache You can disable this option by going to disk manager right click the partition Properti
260. ics section provides measurements about read and write operations in each rank of an extent pool It also shows the number of arrays and the array width of all ranks These values show the current configuration The wider the rank the more performance capability it has By changing these values in your configuration you can influence the throughput of your work Also response and transfer statistics are available with the Postprocessor Cache Activity report generated from SMF record 74 subtype 5 These statistics are provided at the subsystem level in the Cache Subsystem Activity report and at the volume level in the Cache Device Activity report In detail RMF provides the average response time and byte transfer rate per read and write requests These statistics are shown for the I O activity which is called host adapter activity and for the transfer activity from hard disk to cache which is called disk activity Reports were designed for reporting FICON channel utilization RMF also provides support for Remote Mirror and Copy link utilization statistics This support was delivered by APAR OA04877 PTFs are available for the release of z OS V1R4 and later Cache reporting RMF cache reporting and the results of a LISTDATA STATUS command report a cache size that is half the actual size because the information returned represents only the cluster to which the logical control unit is attached Each LSS on the cluster reflects the cache and nonvol
261. iderations 9 5 In the Adapter Settings panel bottom panel of Figure 2 3 for Host Adapter BIOS select Enabled and then press Enter Fast UTIL Options Scan Fibre Devices Ft onf igration Settings L S E Selectable Boot Settings Aurpter Settings BIOS Address BIOS Revision Adapter Serial Number Interrupt Level Adapter Port Nane Host Adapter BIOS Hdapter Port Nane Host Adapter BIOS Frane Size Figure 2 3 Enabling the QLogic HBA BIOS extension 6 In the Adapter Settings panel bottom panel of Figure 2 3 for Host Adapter BIOS select Enabled and then press Enter 7 Inthe Adapter Settings panel press Esc to return to the Configuration Settings panel 8 In the Configuration Settings panel select Selectable BIOS Settings and press Enter 9 In the Selectable BIOS Settings panel press Enter 10 IBM System Storage DS8000 Host Attachment and Interoperability 10 In the Selectable Boot Settings panel modify the boot settings a Change Selectable Boot to Enabled and press Enter as shown in the middle and lower panels of Figure 2 4 onfiguration Settings Adapter Settings Dantann NMafanlt Cattinan Selectable Boot Primary Boot Port Name Lun Boot Port Nane Lun Boot Port Nane Lun Boot Port Name Lun ettings BABAAOAAAAAABAAA BABAHOOOABAABABA BOBOHBBHABAABABA BBBBOOHBABOBBBBA Press C to clear a Boot Port Name en ry Selectable Boot
262. ilabl hdisk15 Availabl hdisk16 Availabl hdisk17 Availabl vpathO Availabl vpathl Availabl vpath2 Availabl vpath3 Availabl 30 T1 01 IBM FC 2107 30 T1 01 IBM FC 2107 30 T1 01 IBM FC 2107 30 T1 01 IBM FC 2107 30 T1 01 IBM FC 2107 30 T1 01 IBM FC 2107 30 T1 01 IBM FC 2107 30 T1 01 IBM FC 2107 31 T1 01 IBM FC 2107 31 T1 01 IBM FC 2107 31 T1 01 IBM FC 2107 31 T1 01 IBM FC 2107 31 T1 01 IBM FC 2107 31 T1 01 IBM FC 2107 31 T1 01 IBM FC 2107 31 T1 01 IBM FC 2107 Data Path Optimizer Pseudo Device Driver Data Path Optimizer Pseudo Device Driver Data Path Optimizer Pseudo Device Driver Data Path Optimizer Pseudo Device Driver ooo 70 DOD DMDMDODDDODMOMO MOM MO MO Management of SDD devices requires more steps in comparison to SDDPCM devices To remove a specific volume from the system remove all hdisks that represent paths to a volume and then the vpath To identify which hdisk identifies which path use the datapath query essmap command or the datapath query device command which provides additional device information In Example 5 17 device information is provided for devices 0 and 1 Example 5 17 datapath query device datapath query device 0 1 DEV 0 DEVICE NAME vpathO TYPE 2107900 POLICY Optimized SERIAL 75TV1812200 Chapter 5 AlX considerations 73 Path Adapter Hard Disk State Mode Select Errors 0 fscsi0 hdisk2 OPEN NORMAL 1297 0 1 fscsi0 hdisk6 OPEN NORMAL 1336 0 2 fscsil hdisk10 OPEN NORMAL
263. ile described in Example 6 17 are based on SLES9 and SLES10 SLES11 uses udev rules that are automatically created by the Yer Another Setup Tool YAST when you use it to determine and configure SAN attached volumes The udev rules can be complicated and are not well documented With SLES11 use YAST for easier configurations 96 IBM System Storage DS8000 Host Attachment and Interoperability The configuration files contain a remote DS8000 port a LUN pair for each path to each volume Example 6 18 demonstrates two DS8000 volumes to the HBA 0 0 0500 going through two separate DS8000 host ports Example 6 18 HBA configuration file Inxmnt01 cat etc sysconfig hardware hwcfg zfcp bus ccw 0 0 0500 bin sh hwcfg zfcp bus ccw 0 0 0500 Configuration for the zfcp adapter at CCW ID 0 0 0500 Configured zfcp disks ZFCP_LUNS 0x500507630a00029f 0x4049400000000000 0x500507630a00029F 0x4049400100000000 0x500507630a0b029f 0x4049400000000000 0x500507630a0b029f 0x4049400100000000 The ZFCP_LUNS statement in the file defines the remote port volume relations or paths that the zfcp driver sets up when it starts The first term in each pair is the WWPN of the DS8000 host port the second term after the colon is the LUN of the DS8000 volume Tip The LUN provided in Example 6 18 is the same LUN in the DS8000 LUN map as shown in Example 6 19 This LUN is padded with zeroes such that it reaches a length of ei
264. imum of eight controller ports to the SVC per controller regardless of the number of nodes in the cluster Configure 16 controller ports for large controller configurations where more than 48 ranks are being presented to the SVC cluster Additionally it is advisable that no more than two ports of each of the DS8000 s 4 port adapters are used Table 13 1 shows the preferred number of DS8000 ports and adapters based on rank count 282 IBM System Storage DS8000 Host Attachment and Interoperability Table 13 1 Preferred number of ports and adapters Ranks Ports Adapter 2 48 8 4 8 gt 48 16 8 16 Configurations gt Configure a minimum of eight ports per DS8000 gt Configure 16 ports per DS8000 when gt 48 ranks are presented to the SVC cluster gt Configure a maximum of two ports per four port DS8000 adapter gt Configure adapters across redundant SAN networks from different I O enclosures 13 3 2 LUN masking For a given storage controller all SVC nodes must see the same set of LUNs from all target ports that have logged into the SVC nodes If target ports are visible to the nodes that do not have the same set of LUNs assigned SVC treats this situation as an error condition and generates error code 625 Validating the LUN masking from the storage controller and then confirming the correct pathcount from within the SVC are critical Example 13 1 shows shostconnect output from the DS8000 Here you can se
265. in the examples in this chapter Copyright IBM Corp 2011 2013 All rights reserved 151 8 1 Available resources For the latest available supported Mac OS X configuration and required software patches see the IBM SSIC at the following website http www ibm com systems support storage ssic interoperability wss For information to prepare the host to attach the DS8000 storage system see the DS8000 Information Center at the following website http publib boulder ibm com infocenter dsichelp ds8000ic index jsp In the Information Center select Configuring Attaching Hosts gt Apple Macintosh host attachment For information regarding the ATTO HBA see the Product Information pages for ATTO Technology Inc at the following vendor website Vendor Website http www attotech com 8 2 Configuring the Apple host on a DS8000 To configure a server on the DS8000 define a host connection using the DS graphical user interface GUI or the DS command line interface CLI On the DS8000 storage system use the predefined host type Apple OS X to automatically configure the volumes in an Mac OS X preferred method For general information see the IBM System Storage DS8000 Host Systems Attachment Guide GC27 2298 02 Tip A restart of the Apple server is required to detect any newly created LUNs The LUN numbering seen in the ATTO Configuration Tool is created by the software and it is set by the order that the LUNs appear du
266. ing factors gt No one person or organization can be held responsible or called for support gt Depending on the target group the distributions differ in the kind of support that is available gt Linux is available for almost all computer architectures gt Linux is rapidly evolving These factors present challenges for generic support of Linux Therefore the product integration support approach by IBM helps limit the uncertainty and the amount of testing IBM supports the following Linux distributions that are targeted at enterprise clients gt Red Hat Enterprise Linux gt SUSE Linux Enterprise Server These distributions have major release cycles of about 18 months These release cycles are maintained for five years and require you to sign a support contract with the distributor They also have a schedule for regular updates These factors help reduce the amount of issues due to the lack of support listed previously Also the limited number of supported distributions enables IBM to work closely with the vendors to ensure interoperability and support Details about the supported Linux distributions can be found in the IBM SSIC matrix at the following website http www ibm com systems support storage ssic interoperability wss 6 1 2 Attaching Linux server to DS8000 resources 82 Use the information provided in the following sections to help attach Linux server to a DS8000 storage subsystem The DS8000 Host Systems
267. inish Disk LUN Ready to Complete Disk layout IBM Fibre Channel Disk naa 600 Non SSD 20 00 GB jumfs devices disks naa 600507630affc29f0000000000002000 GPT i Primary Partitions apacit VMFS IBM Fibre Channel Disk naa 20 00 GB File system Properties Datastore name DS8000_Datastore1 Formatting File system vmfs 5 Block size 1MB Maximum file size 2 00 TB Figure 7 15 Add storage summary page Chapter 7 VMware vSphere considerations 129 9 The new datastore will be listed under Configuration gt Storage If you select the datastore more information is shown in the Datastore Details area Figure 7 16 View Datastores Devices Datastores Refresh Delete Add Storage Rescan All Identification 2 Status Device Drive Type Capacity Free Type Last Update E DSs8000_Datastorel Normal IBM Fibre Channel Non SSD 19 75 GB 18 85 GB VMFSS 31 05 2012 13 52 34 Local_Disk Normal LSILOGIC Serial A Non SD 196 25 GB 64 69 GB VMFSS 31 05 2012 13 52 34 4 m Datastore Details Properties DS8000_Datastore1 19 75GB Capacity Location wmfs volumes 4fc73922 2f50777d 10f2 00215e332F46 4 Hardware Acceleration Unknown 920 00MB W Used 18 85GB Free Refresh Storage Capabilities System Storage Capability N A User defined Storage Capability N A Path Selection Properties Extents Storage I O Control wants atin Volume Label DS8000_Dat
268. integration services are required for the virtual SCSI adapters because they include the drivers for the virtual adapters Microsoft Hyper V server Parent partition with Windows 2008 RIERA Child partition Child partition as lt a Pass through disk D Virtualization layer hypervisor ae Hardware Figure 3 14 Hyper V storage concept 38 IBM System Storage DS8000 Host Attachment and Interoperability 3 10 3 Assigning a VHD to a virtual machine The following steps are required to attach a new VHD to a virtual machine 1 In Hyper V manager select the virtual machine right click and select Settings Hyper Manager a Fie Action view Window Help 4lm Ole 34 Hyper V Manager a a WIN GD307VJHE1IL irtual Machines Name State SLES11 Running Win ae ning Connect Turn Off Shut Down Save Figure 3 15 Virtual machine settings 2 Select the controller to which the new VHD needs to be attached select Hard Drive and click Add B Settings for Windows 2003 A windows 2003 4 Q amp Hardware 4 Add Hardware KE SCSI Controller Ki BIOS You can add hard drives to your SCSI controller or remove the SCSI controller from the Boot from CD machine mE Memory Click on the Add button to add a new hard drive to this SCSI controller 1024 MB a Processo
269. into one logical disk This layer also manages path failover when a path becomes unusable and balances I O requests across the available paths To evaluate the customer configuration to ensure proper performance and installation of the devices see the following SAN Zoning documentation http www ibm com support techdocs atsmastr nsf WebIndex WP101914 2 4 1 Subsystem Device Driver For most operating systems that are supported for DS8000 attachment IBM provides the IBM Subsystem Device Driver SDD at no cost to help support the following functionality Enhanced data availability through automatic path failover and failback Increased performance through dynamic I O load balancing across multiple paths The ability for concurrent download of licensed internal code User configurable path selection policies for the host system vvvy IBM Multipath SDD provides load balancing and enhanced data availability in configurations with more than one I O path between the host server and the DS8000 SDD performs dynamic load balancing across all available preferred paths to ensure full utilization of the SAN and HBA resources You can download SDD from the following website http www ibm com support dlsearch wss rs 540 amp tc ST52G7 amp dc D430 For help with SDD see the IBM System Storage Multipath Subsystem Device Driver User s Guide GC52 1309 It contains all the information that is needed to install configure and use SDD for all supported
270. inux on System z provides a kernel module for DASD It creates a fixed block address range within a data set that resides on a CKD device This emulated FB storage appears to the Linux system as though it where a SCSI disk For redundancy DASDs usually are connected through more than one channel The multipathing is managed in the I O subsystem of the System z machine Therefore the Linux system only sees one device Linux also supports parallel access volume PAV with DASDs This increases the throughput by enabling several I O requests to the same device in parallel such as the System z method of command queuing In recent Linux distributions SLES11 Red Hat Enterprise Linux 6 dynamic PAV is supported and managed by the DASD driver completely without user intervention In older Linux versions only static PAV is supported and must be configured manually Additional details about PAV for Linux on System z from SHARE Inc can be found at the following website http share confex com share 115 webprogram Session7154 html gt FCP attached SCSI devices The FICON channel adapter cards in a System z machine can also operate in Fibre Channel Protocol FCP mode FCP is the protocol that transports SCSI commands over the Fibre Channel interface It is used in all open systems implementations for SAN attached storage Certain operating systems that run on a System z mainframe can use this FCP capability and connect directly to FB storage devices
271. ion in either native FICON or ZHPF protocols in response to the type of request being performed These are nondisruptive changes For z OS after the PTFs are installed in the LPAR you must then set ZHPF YES in IECIOSxx in SYS1 PARMLIB or issue the SETIOS ZHPF YES command The default setting is ZHPF NO Use the ZHPF YES setting after the required configuration changes and prerequisites are met For more information about ZHPF see this website http www ibm com support techdocs atsmastr nsf fe582a1e48331b5585256de50062aelc 05d19b2a9bd95d4e8625754c0007d365 0penDocument 12 5 6 ZHPF latest enhancements Starting with release level 6 2 LMC code level 7 6 20 xx DS8000 supports a wide range of new enhancements specific for ZHPF channels The following sections provide a list of the enhancements with a brief description of each Enhancements for QSAM BSAM The typical command chain in a command information unit is as follows gt Prefix Read Read Locate Record Read Read gt Prefix Write Write Locate Record Write Write Original ZHPF supported the following read and write commands gt Read Track Data gt Write Track Data ZHPF command support is expanded to include the following options Format Write Update Key Update Data Read Count Read Data Read Key and Data Read Count Key and Data vvvvvvy Supported on z OS V1R11 V1R12 V1R13 and above ZHPF incorrect transfer length support Read a
272. iority Manager can be enabled by web GUI This page displays the properties for the DS8k05_3Tier storage image Make the appropriate modifications and then click OK or Apply to save your changes n General Hardware Advanced MTS Emulated MTS IBM 2107 75ACV21 IBM 2421 75ACV21 WWNN Firmware level 5005076303FFD18E 7 6 30 160 Easy Tier I O Priority Manager Manage gt aes F Send SNMP Traps All Pools Monitor Auto mode is managing No Pools Disable Monitor Auto Mode Volumes All Volumes No Volumes Figure 12 6 Easy Tier and I O Priority Manager setup Tip Easy Tier should work in auto mode by selecting the All pools option to obtain the maximum performance Chapter 12 IBM System z considerations 277 12 9 TPC R V5 1 in a z OS environment The IBM Tivoli Storage Productivity Center for Replication for System z products are designed to support hundreds of replication sessions across thousands of volumes for both open and System z attached volumes These products starting at V5 1 have an intuitive graphical user interface GUI and also a command line interface CLI to help you to configure pairing more easily For more information about system requirements and IBM HyperSwap activity see following the Redbooks publications gt IBM Tivoli Storage Productivity Center for Replication for System z V5 1 SC27 4055 gt IBM Tivoli Storage Productivity Center for Replication f
273. ioscan fnC disk Class I H W Path Driver S W State H W Type Description USB SCSI Stack Adaptor dev deviceFileSystem Usb MassStorage dsk disk hp 1008 294 A60020000001 disk 1 dev deviceFileSystem Usb MassStorage rdsk disk hp 1008 294 A60020000001 0 0 2 1 0 16 UsbScsiAdaptor CLAIMED DEVICE disk 6 0 3 1 0 1 12 255 0 0 0 sdisk CLAIMED DEVICE dev dsk c2t0d0 dev rdsk c2t0d0 disk 7 0 3 1 0 1 12 255 0 0 1 sdisk CLAIMED DEVICE dev dsk c2t0d1 dev rdsk c2t0d1 disk 8 0 3 1 0 1 13 255 0 0 0 sdisk CLAIMED DEVICE dev dsk c3t0d0 dev rdsk c3t0d0 disk 9 0 3 1 0 1 13 255 0 0 1 sdisk CLAIMED DEVICE dev dsk c3t0d1 dev rdsk c3t0d1 disk 4 0 3 1 0 1 14 255 0 0 0 sdisk CLAIMED DEVICE dev dsk c1t0d0 dev rdsk c1t0d0 disk 5 0 3 1 0 1 14 255 0 0 1 sdisk CLAIMED DEVICE dev dsk clt0d1 dev rdsk c1t0d1 disk 2 0 3 1 0 1 15 255 0 0 0 sdisk CLAIMED DEVICE dev dsk cOt0d0 dev rdsk cOt0d0 disk 3 0 3 1 0 1 15 255 0 0 1 sdisk CLAIMED DEVICE dev dsk cOt0d1 dev rdsk cOt0d1 disk 10 0 4 1 0 0 0 0 0 sdisk CLAIMED DEVICE dev dsk c30t0d0 dev dsk c30t0d0s2 dev rdsk c30t0d0 dev dsk c30t0d0s1 dev dsk c30t0d0s3 dev rdsk c30t0d0s1 disk 11 0 4 1 0 0 0 1 0 sdisk CLAIMED DEVICE dev dsk c30t1d0 dev rdsk c30t1d0 IBM 2107900 IBM 2107900 IBM 2107900 IBM 2107900 IBM 2107900 IBM 2107900 IBM 2107900 IBM 2107900 HP DHO72ABAA6 dev rdsk c30t0d0s2 dev rdsk c30t0d0s3 HP DHO72ABAA6 The ioscan Nm command also shows the relationship bet
274. ipathd O off l off 2 o0ff 3 on 4 off 5 on 6 off 100 IBM System Storage DS8000 Host Attachment and Interoperability Enabling multipathing for Red Hat Enterprise Linux 6 There are some new features that are related to the DS8000 storage in Red Hat 6 There are no major differences in setting up the Red Hat 6 from the set up outline in 5 However a good practice for this setup is to follow Chapter 2 Storage Consideration in the documentation from the official Red Hat 6 website http docs redhat com docs en US Red_Hat_Enterprise_Linux 6 pdf Storage_Administr ation_Guide Red_Hat_Enterprise_Linux 6 Storage_Administration_Guide en US pdf It is important to have previously installed the device mapper multipath package 32 or 64 version from the Red Hat Enterprise 6 CD Create the storage volumes and volume groups and create the multipath conf file as stated in the Storage Administration Guide mentioned Tip When you create the multipath conf file under the etc directory make sure to copy and paste the script from the RHEL 6 Storage Administration Guide without spaces because it can give a keyboard error when starting the multipath server See Figure 6 2 on page 98 for an example of a multipath set up Red Hat 6 has new features that can improve workloads with the DS8000 storage units gt File System Cashing This feature allows the use of local storage data for cash over the network which can lead to decreased traffic network
275. irtual HBAs Example 6 2 Volumes mapped through NPIV virtual HBAs p6 570 lparl3 1sscsi 1 0 0 0 disk IBM 2107900 3 23 dev sdc 1 0 0 1 disk IBM 2107900 3 23 dev sdd 1 0 0 2 disk IBM 2107900 3 23 dev sde 2 0 0 0 disk IBM 2107900 3 23 dev sdm 2 0 0 1 disk IBM 2107900 3 23 dev sdn 2 0 0 2 disk IBM 2107900 3 23 dev sdo To maintain redundancy use more than one virtual HBA each one running on a separate real HBA In this case DS8000 volumes will show up once per path and will need to be managed by a DM Multipath Chapter 6 Linux considerations 87 System z For Linux running on an IBM System z server there are even more storage attachment choices and therefore potential confusion Linux running natively in a System z LPAR To run Linux directly on a System z LPAR two methods can be used to attach disk storage gt IBM FICON channel attached count key data CKD devices Linux on System z supports channel attached CKD devices These type of devices are used in traditional mainframe environments like z OS and are called direct access storage devices DASDs DASDs are connected to the Linux LPAR using the regular System z methods with the I O definition file IODF and hardware configuration definition HCD By design Linux cannot use CKD devices Its I O access methods are all designed for fixed block FB storage such as SCSI attached disks as it is used in all open systems implementations To overcome this L
276. irtual disks between VMs or to assign an existing virtual disk 5 Define the properties of the new virtual disk as shown in Figure 7 29 a Disk Size Size of the new virtual disk It is the size that the operating system will see b Disk Provisioning e Thick Provision Lazy Zeroed The full capacity is allocated on the datastore on creation The blocks are not overwritten by zeros the VM might be able to read any data that was previously stored on these blocks The virtual disk is available almost instantly e Thick Provision Eager Zeroed The full capacity is allocated on the datastore on creation The blocks are all overwritten by zeros to delete anything that was previously stored on these blocks This method requires the most time but ensures that a VM cannot read any old information from the blocks e Thin Provision The capacity is allocated on demand when the VMs operating system tries to access it This method is saving physical space on the datastore and allows for overcommitment 136 IBM System Storage DS8000 Host Attachment and Interoperability c Storage location of the new virtual disk e Store with the virtual machine The new virtual disk will be stored on the same datastore as the rest of the virtual machine e Specify a datastore The new virtual disk will be stored on a different datastore than the rest of the virtual machine Possible reasons could be that the original datastore does not have enough cap
277. is activation process might take time to complete on all attached disk storage subsystems that support HyperPAV Verify that HyperPAV is active through a subsequent display command as shown in Example 12 7 Example 12 7 Activate HyperPAV dynamically SETIOS HYPERPAV YES I0S189I HYPERPAV MODE CHANGE INITIATED CONTROL UNIT CONVERSION WILL COMPLETE ASYNCHRONOUSLY D I0S HYPERPAV IOS098I 15 55 06 HYPERPAV DATA 457 HYPERPAV MODE IS SET TO YES If you are currently using PAV and FICON then no HCD or DS8000 logical configuration changes are needed on the existing LSSs To stage HyperPAV development follow these steps 1 Load or authorize the HyperPAV feature on the DS8000 If necessary you can run without exploiting this feature by using the z OS PARMLIB option 2 Enable the HyperPAV feature on z OS images that you want to use HyperPAV using the PARMLIB option or the SETIOS command 3 Enable the HyperPAV feature on all z OS images in the sysplex and authorize the licensed function on all attached DS8000s 4 Optional Reduce the number of aliases defined Full coexistence with traditional PAVs such as static PAV or dynamic PAV and sharing with z OS images without HyperPAV enabled helps migration to HyperPAV to be a flexible procedure Chapter 12 IBM System z considerations 265 HyperPAV definition The correct number of aliases for your workload can be determined from the analysis of RMF data The PAV analysis tool which c
278. is provided in the following sections The Linux distributors documentation pages are good starting points for installation configuration and administration of Linux servers especially when working with server platform specific issues This documentation can be found at the following websites gt Novell SUSE Linux Enterprise Server http www novell com documentation suse html gt Red Hat Enterprise Linux http www redhat com docs manuals enterprise For more information about storage attachment using FCP see the System z I O Connectivity page at the following website http www ibm com systems z connectivity products The IBM System z Connectivity Handbook SG24 5444 contains information about connectivity options available for use within and beyond the data center for IBM System z servers This handbook has a section for Fibre Channel attachment You can access this book at the following website http www redbooks ibm com redbooks nsf RedbookAbstracts sg245444 html 84 IBM System Storage DS8000 Host Attachment and Interoperability 6 1 3 Understanding storage related improvements to Linux This section provides a summary of storage related improvements that were introduced to Linux in recent years and details about usage and configuration Limitations and issues of older Linux versions The following list provides limitations and issues documented with older Linux versions These limitations are not valid with newer
279. ity 5 3 1 Queue depth tuning Figure 5 3 represents the order in which software comes into play over time as the I Os traverse the stack Application File System optional LVM device driver optional SDD SDDPCM or other Physical Disk Figure 5 3 1 O stack from the application to the disk The reason to submit more than one I O to a disk is performance Only submitting a single I O at a time can give very good I O service time but very poor throughput The IOPS for a disk is limited by queue_depth average I O service time The queue depth is the number of I O operations that can run in parallel on a device The service time is the time needed to physically process the I O Assuming a queue_depth of 3 and an average I O service time of 10 ms this yields a maximum throughput of 300 IOPS for the hdisk And for many applications it might not be enough throughput As I Os traverse the I O stack AIX needs to keep track of them at each layer So I Os are essentially queued at each layer Generally some number of in flight I Os can be issued at each layer and if the number of I O requests exceeds that number they reside in a wait queue until the required resource becomes available So there is essentially an in process queue and a wait queue at each layer SDD and SDDPCM are a little more complicated At the file system layer file system buffers limit the maximum number of i
280. ize Tbytes Max LV Size Tbytes Max PV Size Tbytes Max VGs Max LVs Max PVs Max Mirrors Max Stripes Max Stripe Size Kbytes Max LXs per LV Max PXs per PV Max Extent Size Mbytes VG Version Max VG Size Tbytes Max LV Size Tbytes Max PV Size Tbytes Max VGs Max LVs Max PVs Max Mirrors Max Stripes Max Stripe Size Kbytes Max LXs per LV Max PXs per PV Max Extent Size Mbytes 1 0 510 16 2 256 255 255 2 255 32768 65535 65535 256 2 0 2048 256 16 512 511 511 5 511 262144 33554432 16777216 256 2 1 2048 256 16 2048 2047 2048 5 511 262144 33554432 16777216 256 198 IBM System Storage DS8000 Host Attachment and Interoperability To create a volume group for large LUNs complete the following steps 1 Assign a large LUN for example 17 TB to HP UX and initialize the LUN by entering the pvcreate command as shown in Example 10 19 The operating system will communicate that the traditional volume group version 1 0 cannot handle large LUNs Define a group special file using the mknod command specifying a major number such as 128 instead of 64 Create the volume group by entering the vgcreate command using the V option for the Version 2 1 volume group Example 10 19 Preparing a large LUN pvcreate f dev rdisk disk1078 pvcreate Warning The physical volume dev rdisk disk1078 is initialized to use only the first 2147483647
281. ks Chapter 4 Virtual I O Server considerations 49 Virtual SCSI is based on a client server relationship as shown in Figure 4 2 The VIOS owns the physical resources and the virtual SCSI server adapter and acts as a server or SCSI target device The client logical partitions have a SCSI initiator referred to as the virtual SCSI client adapter and access the virtual SCSI targets as standard SCSI LUNs The following SCSI peripheral device types are generally supported by the VIOS as backing devices for VSCSI emulation Disk backed by logical volume Disk backed by physical volume Disk backed by file Optical CD ROM DVD RAM and DVD ROM Optical DVD RAM backed by file Tape devices with virtual tape support that enables serial sharing of selected SAS tape devices vvvvvyy Figure 4 2 shows virtual SCSI emulation on a VIOS server Virtual O Server ii I 1 I POWER Hypervisor I A physical HBAs and storage m pr Ol eae gt is VSCSI server adapters Tu C VSCSI client adapters Figure 4 2 VSCSI emulation on VIOS When using virtual SCSI emulation on the VIOS with DS8000 storage system volumes all the DS8000 volumes that are to be used as backing devices for the client partitions need to be mapped directly to the VIOS partition using the WWPNs of the physical Fibre Channel adapters in the VIOS partition You can find the WWPNs of the VIOS Fibre Channel adapters using the 1sdev dev fcsX
282. kup and restore work immediately in an application LPAR with virtual Fibre Channel and NPIV With virtual Fibre Channel physical to virtual device compatibility issues do not arise and even SCSI 2 reserve or release and SCSI 3 persistent reserve methodologies in clustered or distributed environments are available The specific multipathing software such as SDDPCM providing a load balancing path selection algorithm for the mapped SAN devices must be installed directly on the client LPAR not on the VIOS The VIOS partition only acts as a pass through module The LUNs mapped to WWPNs using NPIV are not actually seen by the VIOS With NPIV you can configure the managed system so that multiple logical partitions can access independent physical storage through the same physical Fibre Channel adapter each with their own unique WWPNs The VIOS partition owns the physical Fibre Channel adapters which must support NPIV such as the IBM 8 Gbps PCI Express Dual Port Fibre Channel Adapter The VIOS partition acts as pass through module to bridge Fibre Channel traffic from the configured virtual Fibre Channel host adapters vfchost to the physical adapter ports IBM System Storage DS8000 Host Attachment and Interoperability The client partition owns a virtual Fibre Channel client adapter and directly connects to a virtual Fibre Channel host adapter on the VIO partition as shown in Figure 4 5 Client logical Client logical Client
283. l ASP and loss of access to any disk will cause the system to fail Even with user ASPs loss of a user ASP disk eventually causes the system to stop Independent auxiliary storage pools IASPs provide isolation so that loss of disks in the IASP only affects users that are accessing that IASP but the rest of the system is unaffected However with multipath even loss of a path to a disk in an IASP does not cause an outage With the combination of multipath and RAID 5 RAID 6 or RAID 10 protection in the DS8000 you can take advantage of full protection of the data paths and the data itself without the requirement for additional disks Besides providing resiliency multipath improves performance because IBM i uses all available paths to a LUN in DS8000 The I O operations are balanced among the available paths in a round robin algorithm 11 5 1 Avoiding single points of failure When implementing multipath provide as much redundancy as possible With native attachment to the of DS8000 multipath requires two Fibre Channel IOAs at a minimum when connecting the same logical volumes With IOP less adapters connect the LUNs to two ports from separate adapters Ideally place them on separate buses or loops and in separate I O racks in the Power System If a SAN is included use separate switches for each path Also use host adapters in separate I O drawer pairs in the DS8000 When connecting DS8000 to IBM i with VIOS implement multipath so t
284. l disk g Virtualization layer vmkernel VMFS1 virt disk file VMFS datastore logical volume Figure 7 2 Storage concepts of ESXi Chapter 7 VMware vSphere considerations 125 Creating a VMFS datastore The following steps describe the tasks required to create a datastore on a DS8000 volume 1 Ensure that the host is correctly zoned to the DS8000 see 1 3 SAN considerations on page 3 for details on the SAN configuration Tip To find out the WWPN of the HBA in ESXi select the host click the Configuration tab and select Storage Adapters The WWNN and WWPN are listed in a column next to the adapter name See Figure 7 3 Storage Adapters Device Type 82801G ICH7 Family IDE Controller vmhbad Block SCSI _ vmhba32 Block SCSI ISP81xx based 10 GbE FCoE to PCI Express CNA _ vmhbas Fibre Channel 20 00 00 c0 dd 18 3b 79 21 00 00 c0 dd 18 3b 79 _ vmhba7 Fibre Channel 20 00 00 c0 dd 18 3b 7b 21 00 00 c0 dd 18 3b 7b ISP2532 based 8Gb Fibre Channel to PCI Express HBA _ vmhba2 Fibre Channel 20 00 00 24 ff 2d c3 40 21 00 00 24 ff 2d c3 40 i ymhba3 Fibre Channel 20 00 00 24 ff 2d c3 41 21 00 00 24 ff 2d c3 41 eS vmhba4 Fibre Channel 20 00 00 24 ff 2d c5 68 21 00 00 24 ff 2d c5 68 _ vmhbas Fibre Channel 20 00 00 24 ff 2d c5 69 21 00 00 24 ff 2d c5 69
285. latform sun4us boot_archive 5 Restart your system After restarting your system you can see that Solaris was booted off a DS8000 LUN with MPxIO enabled as shown in Example 9 28 Example 9 28 Solaris that was booted off a DS8000 LUN MPxIO is enabled bash 3 00 df dev dsk c5t600507630AFFC29F0000000000003101d0s0 3667056 blocks 564060 files devices devices 0 blocks 0 files system contract ctfs 0 blocks 2147483621 files proc proc 0 blocks 29932 files etc mnttab mnttab 0 blocks 0 files etc svc volatile swap 6897584 blocks 558165 files system object objfs 0 blocks 2147483465 files etc dfs sharetab sharefs 0 blocks 2147483646 files dev fd fd 0 blocks 0 files tmp swap 6897584 blocks 558165 files var run swap 6897584 blocks 558165 files export home dev dsk c5t600507630AFFC29F0000000000003101d0s7 3263196 blocks 407676 files bash 3 00 format Searching for disks done 178 IBM System Storage DS8000 Host Attachment and Interoperability AVAILABLE DISK SELECTIONS 0 cOtOd0 lt FUJITSU MAW3073NC 3701 cyl 24345 alt 2 hd 8 sec 737 gt pci 83 4000 FJSV ulsa 2 1 sde 0 0 1 clt0d0 lt FUJITSU MAW3073NC 3701 cyl 24345 alt 2 hd 8 sec 737 gt pci 83 4000 FISV ulsa 2 sd 0 0 2 c5t600507630AFFC29F0000000000003100d0 lt IBM 2107900 268 cyl 4367 alt 2 hd 30 sec 64 gt scsi_vhci ssd g600507630af fc29f0000000000003100 3 5t600507630AFFC29F0000000000003101d0 lt IBM 2107900 268 cyl 8736 alt
286. le systems are now reported as residing on dev dsk c4t500507630A00029Fd1s0 and dev dsk c4t500507630A00029Fd1s7 The DS8000 LUN selected when installing Solaris from CDROM is c4t500507630A00029Fd1 MPxIO was disabled and you can see twice as many DS8000 LUNs each separate path is reported as shown in Example 9 25 Example 9 25 Solaris that was booted off a DS8000 LUN with MPxIO disabled bash 3 00 df dev dsk c4t500507630A00029Fd1s0 3667182 blocks 564064 files devices devices 0 blocks 0 files system contract ctfs 0 blocks 2147483622 files proc proc 0 blocks 29934 files etc mnttab mnttab 0 blocks 0 files etc svc volatile swap 6897264 blocks 558173 files system object objfs 0 blocks 2147483465 files etc dfs sharetab sharefs 0 blocks 2147483646 files dev fd fd 0 blocks 0 files tmp swap 6897264 blocks 558173 files var run swap 6897264 blocks 558173 files export home dev dsk c4t500507630A00029Fd1s7 3263196 blocks 407676 files bash 3 00 format 176 IBM System Storage DS8000 Host Attachment and Interoperability Searching for disks done AVAILABLE DISK SELECTIONS 0 cOt0d0 lt FUJITSU MAW3073NC 3701 cyl 24345 alt 2 hd 8 sec 737 gt pci 83 4000 FUSV ulsa 2 1 sd 0 0 1 cltQd0 lt FUJITSU MAW3073NC 3701 cyl 24345 alt 2 hd 8 sec 737 gt pci 83 4000 FISV ulsa 2 sd 0 0 2 3t500507630A03029Fd0 lt IBM 2107900 268 cyl 4367 alt 2 hd 30 sec 64 gt pci 80 2000 em1 x
287. lity Tip In order to obtain the most accurate results it is highly advisable to select as much detail as possible in the interoperability website DS8000 Host Systems Attachment Guide For information about how to attach an open systems host to your DS8000 storage system see the BM System Storage DS8000 Host Systems Attachment Guide SC26 7917 at the following website http www ibm com support docview wss uid ssg1S7001161 Tip The DS8000 Host Attachment and Interoperability book includes information for the most widely used host Make sure to use the DS8000 host Systems Attachment Guide book to review the full list of supported host by the DS8000 systems Installation and troubleshooting documentation For general installation and troubleshooting documentation see the IBM Support Portal site http www ibm com support us en 2 1 2 HBA vendor resources All of the Fibre Channel HBA vendors have websites that provide information about their products facts and features and support information These sites can be helpful when you need details that cannot be supplied by IBM resources for example when troubleshooting an HBA driver IBM cannot be held responsible for the content of these sites QLogic Corporation See the QLogic website at the following website http www qlogic com QLogic maintains a page that lists all the HBAs drivers and firmware versions that are supported for attachment to IBM storage systems which you
288. logical partition 1 partition 2 partition 3 Client virtual al Client virtual Client virtual fibre channel fibre channel fibre channel adapter ad adapter adapter Virtual I O server Physical fibre channel adapter Hypervisor Server virtual fibre channel adapter Server virtual fibre channel adapter Server virtual fibre channel adapter Storage Area Network Figure 4 5 Virtual Fibre Channel on a single VIOS managed system Each virtual Fibre Channel server adapter on the VIOS partition connects to one virtual Fibre Channel client adapter on a virtual I O client partition Using their unique WWPNs and the virtual Fibre Channel connections to the physical Fibre Channel adapter the operating systems that runs in the client logical partitions discover initialize and manage their physical storage located in the SAN The WWPN for the client LPAR is automatically generated by the system when the virtual Fibre Channel client adapter is created for an active partition The Hardware Management Console HMC generates WWPNs based on the range of names available for use with the prefix from the vital product data within the managed system When a virtual Fibre Channel adapter is removed from a client logical partition the hypervisor deletes the WWPNs that are assigned to the virtual Fibre Channel adapter on the client logical partition Because the HMC does not reuse deleted WWPNs when generating f
289. lowing people for their contributions to this project Bjorn Wesselbaum Stefen Deierling Hans Paul Drum Wilhelm Gardt Torsten Rothenwalat Edgar Strubel IBM xii IBM System Storage DS8000 Host Attachment and Interoperability Now you can become a published author too Here s an opportunity to spotlight your skills grow your career and become a published author all at the same time Join an ITSO residency project and help write a book in your area of expertise while honing your experience using leading edge technologies Your efforts will help to increase product acceptance and client satisfaction as you expand your network of technical contacts and relationships Residencies run from two to six weeks in length and you can participate either in person or as a remote resident working from your home base For more information about the residency program browse the residency index or apply online see the following website ibm com redbooks residencies html Comments welcome Your comments are important to us We want our books to be as helpful as possible Send us your comments about this book or other IBM Redbooks publications in one of the following ways gt Use the online Contact us review Redbooks form found at this website ibm com redbooks gt Send your comments in an email to redbooks us ibm com gt Mail your comments to IBM Corporation International Technical Support Organization Dept HYTD Mail Station P0
290. lps provide better control of performance of IBM i workload However you might consider sharing ranks when the other servers workloads have a sustained low disk I O rate compared to the System i I O rate Generally System i has a relatively high I O rate where that of other servers might be lower often under one I O per GBps As an example a Windows file server with a large data capacity can typically have a low I O rate with fewer peaks and can be shared with IBM i ranks However Microsoft SQL Server IBM DB2 Server or other application servers might show higher rates with peaks therefore consider using other ranks for these servers 216 IBM System Storage DS8000 Host Attachment and Interoperability When sharing DS8000 among many IBM i workloads dedicate the ranks to the most important IBM i systems to help provide stable performance Also consider sharing the ranks among less important or smaller IBM i workloads However the decision to mix platforms or mix IBM i workloads on a DS8000 array rank or extent pool is based on your IBM i performance requirements 11 4 3 Connecting using SAN switches When connecting DS8000 series to IBM i using switches consider the following points for your SAN configuration gt Zone the SAN switches so that one zone contains one or multiple IBM i ports and one DS8000 port The reason for this advice is that IBM i establishes only one path from an IBM i port to a DS8000 port even if IBM i
291. ls of your selection Furthermore a list of supported storage area network SAN switches and directors is displayed Alternatively you can generate and download a Microsoft Excel data file that contains information based on the product family model family and product version that you have selected The information in this file is similar to the information that was in the previous interoperability matrix for System Storage Select the family model and version of your product and click the Export Selected Product Version xls link as shown in Figure 2 1 to generate an Excel data file based on your selections System Storage Interoperation Center SSIC SSIC Education and Help Please view the details of the interoperability configurations queried This requires exporting the data or clicking the Submit button at the bottom of the search interface then clicking on the details link in the results table Revise Selected Criteria click link below to change search query 1 Product Family 2 Product Model 3 Product Version New Search Configuration Results 14590197 Product Family Product Model Product Version DS8100 DS8300 R3 1 bundle 63 10_xx DS8100 DS8300 R4 bundle 64 00_xx DS8100 DS8300 R4 21 bundle 64 21 xx Export Selected Product Version xls Figure 2 1 Generating an Excel data file based on your selections 6 IBM System Storage DS8000 Host Attachment and Interoperabi
292. lt y 192 IBM System Storage DS8000 Host Attachment and Interoperability Add disks as spare disks for dg01 y n q default n n Exclude disks from hot relocation use y n q default n A new disk group will be created named dg01 and the selected disks will be added to the disk group with default disk names cl0t6d0 cl0t6d1 Continue with operation y n q default y Do you want to use the default layout for all disks being initialized y n q default y n Do you want to use the same layout for all disks being initialized y n q default y Enter the desired format cdsdisk hpdisk q default cdsdisk hpdisk Enter the desired format cdsdisk hpdisk q default cdsdisk hpdisk Enter desired private region length lt privlen gt q default 1024 Initializing device cl0t6d0 Initializing device cl0t6dl VxVM NOTICE V 5 2 120 Creating a new disk group named dg01 containing the disk device cl0t6d0 with the name dg0101 VxVM NOTICE V 5 2 88 Adding disk device cl0t6d1 to disk group dg01 with disk name dg0102 Add or initialize other disks y n q default n n vxdisk list DEVICE TYPE DISK GROUP STATUS c2t0d0 auto none online invalid c2t1d0 auto none online invalid cl0t0d1 auto none online invalid cl10t6d0 auto hpdisk dg0101 dg01 online cl0t6d1 auto hpdisk dg0102 dg01 online cl0t6d2 auto none online invalid Chapter 10 HP UX considerations 193 The graphical equival
293. lume to a volume group Using DS Storage Manager GUI When assigning a volume to a volume group follow these steps 1 Go to Volumes Volume Groups select volume group Action Properties 2 From the Action drop down men select Add volumes Select the volume and click Next 3 In the LUN ID assignment window double click the value in the LUN ID column and enter the desired value as in Figure 7 17 Volume Groups eee LUN ID assignments Volume Group Nickname ITSO ESX 1 Review the LUN ID assignments and verify that they are correct You may change the assignments by double clicking on the value you wish to change and typing in the new value Only integers between 0 and 255 will be accepted Clicl Next when done with the assignments B 4 M Extent Pool Showing 1 items Figure 7 17 Specify LUN ID 4 Click Next and Finish Using DS CLI When assigning a volume to the volume group with chvolgrp use the parameter 1un to specify the LUN ID as shown in Example 7 2 Example 7 2 Specify LUN ID dscli gt chvolgrp action add volume A005 lun 1 v64 Date Time 1 Juni 2012 13 11 08 CEST IBM DSCLI Version 7 6 20 221 DS IBM 2107 75TV181 CMUCO0031I chvolgrp Volume group V64 successfully modified dscli gt showvolgrp lunmap v64 Date Time 1 Juni 2012 13 11 12 CEST IBM DSCLI Version 7 6 20 221 DS IBM 2107 75TV181 Name ITSO ESX 1 ID V64 Type SCSI Map 256 Vols A005
294. m dev vx dsk 20781_dg 20781_vol as shown in Example 9 19 Example 9 19 df command after file system expansion bash 3 00 df k Filesystem kBytes used avail capacity Mounted on dev dsk c1t0d0s0 14112721 4456749 9514845 32 devices 0 0 0 0 devices ctfs 0 0 0 0 system contract proc 0 0 0 0 proc mnttab 0 0 0 0 etc mnttab swap 7222640 1160 7221480 1 etc svc volatile objfs 0 0 0 0 system object fd 0 0 0 0 dev fd swap 7221480 0 7221480 0 tmp swap 7221528 48 7221480 1 var run swap 7221480 0 7221480 0 dev vx dmp swap 7221480 0 7221480 0 dev vx rdmp dev dsk c1lt0d0s7 20160418 2061226 17897588 11 export home dev vx dsk 03461_dg 03461_vol 10485760 20062 9811599 1 03461 dev vx dsk 20781_dg 20781_vol 17825792 21861 16691193 1 20781 Chapter 9 Solaris considerations 171 9 6 Booting from SAN Solaris can be installed and booted off a SAN volume as opposed to a server internal disk This section provides information about how to boot Solaris from a DS8000 LUN To install an operating system onto a LUN that resides in the SAN the server needs to have access to it It means that the Fibre Channel adapters in the server need to have a facility to log in to a SAN fabric and scan and provision a LUN as a disk available to the operating system installation routine This facility is provided by special boot code that needs to be loaded into the Fibre Channel adapters For this load you need a running operati
295. makes specified volumes available when it starts Volumes and their path information the local HBA and DS8000 ports are defined in configuration files Tip SLES10 SP3 was used for the production of the examples in this section Procedures commands and configuration files of other distributions can differ Linux on System z was used to produce Example 6 15 with two FC HBAs assigned through z VM In z VM you can determine the device numbers of these adapters Example 6 15 FCP HBA device numbers in z VM CP QUERY VIRTUAL FCP FCP 0500 ON FCP 5A00 CHPID 8A SUBCHANNEL 0000 FCP 0600 ON FCP 5B00 CHPID 91 SUBCHANNEL 0001 The Linux on System z tool to list the FC HBAs is 1szfcp It shows the enabled adapters only Adapters that are not listed correctly can be enabled using the chccwdev command as shown in Example 6 16 Example 6 16 List and enable the Linux on System z FCP adapters lnxmnt01 lszfcp 0 0 0500 host0 Inxmnt01 chccwdev e 600 Setting device 0 0 0600 online Done Inxmnt01 Iszfcp 0 0 0500 host0 0 0 0600 host1 For SUSE Linux Enterprise Server 10 the volume configuration files reside in the etc sysconfig hardware directory There must be one configuration file for each HBA Example 6 17 provides their naming scheme Example 6 17 HBA configuration files Inxmnt01 1s etc sysconfig hardware grep zfcp hwcfg zfcp bus ccw 0 0 0500 hwcfg zfcp bus ccw 0 0 0600 Using YAST The type of configuration f
296. me for the new VHD and specify where to store it Click Next See Figure 3 19 Specify the name and location of the virtual hard disk file Name datadisk vhd Location Ewen Browse Figure 3 19 Filename and location for VHD 6 Specify the size of the VHD or choose an existing physical disk from which to clone the VHD See Figure 3 20 You can create a blank virtual hard disk or copy the contents of an existing physical disk Create a new blank virtual hard disk Size 10 GB Maximum 2040 GB Copy the contents of the specified physical disk Physical Hard Disk sze PHYSICALDRIVEO 556 GB PHYSICALDRIVE1 199 GB PHYSICALDRIVES 19GB PHYSICALDRIVE4 49 GB Figure 3 20 VHD size Tip To virtualize a physical host it is possible to create the new VHD as a copy of an existing physical disk by selecting Copy the contents 7 On the summary page click Finish Attention If a VHD is removed from a virtual machine the VHD file is not automatically deleted To free up the space occupied delete it manually from the partition As illustrated in Figure 3 13 on page 37 the Windows Server 2008 used for management of the Hyper V setup is also running a virtual machine Therefore it is also possible to attach a VHD to it Follow these steps 1 Inside the disk management right click More Actions and select Create VHD See Figure 3 21 Disk Management Mo
297. ment suite to manage multiple ESXi hosts and to provide advanced functions such as on line migrations cloning high availability and so on gt vSphere Client A Windows software used to connect to either single ESXi hosts or vCenter Server that provides a graphical user interface vCenter Server virt layer virt layer vSphere Client Figure 7 1 vSphere components 124 IBM System Storage DS8000 Host Attachment and Interoperability 7 2 vSphere storage concepts This section focuses on the concepts of how storage is handled by ESXi 7 2 1 VMFS ESXi uses a special file system to store virtual machines called Virtual Machine File System VMFS VMFS is a clustered file system that can be accessed concurrently by up to 8 ESXi hosts A VMFS partition can be spread over 32 logical volumes called extents It can have a maximum size of 64 TB and the maximum file size on a VMFS partition is 2 TB On a VMFS partition ESXi stores all the files of which a VM consists Virtual machine configuration file vmx Virtual machine BIOS file nvram Virtual disk files or raw device mappings vmdk Virtual machine log files log vvvy VMs can use two different concepts for their virtual harddrives virtual disks and raw device mappings RDM Both concepts are shown in Figure 7 2 and explained further in this chapter ESXi Server virtual Windows VM Linux VM Windows VM machine RDM virtua
298. moving IOPs and multipath connections between nodes Do not split multipath connections between nodes either by moving IOPs between logical partitions LPAR or by switching expansion units between systems If two separate nodes have both connections to the same LUN in the DS8000 both nodes might overwrite data from the other node Chapter 11 IBM i considerations 219 The multiple system environment enforces the following rules when using multipath disk units gt If you move an IOP with a multipath connection to a separate LPAR you must also move all other IOPs with connections to the same disk unit to the same LPAR gt When you make an expansion unit a switchable entity make sure that all multipath connections to a disk unit switch with the expansion unit gt When you configure a switchable entity for a independent disk pool make sure that all of the required IOPs for multipath disk units switch with the independent disk pool If a multipath configuration rule is violated the system issues warnings or errors to alert you about the condition It is important to pay attention when disk unit connections are reported missing You want to prevent a situation where a node might overwrite data on a LUN that belongs to another node Disk unit connections might be missing for a variety of reasons but especially if one of the preceding rules was violated If a connection for a multipath disk unit in any disk pool is found to be missing du
299. mple 4 2 shows the output when you create a host attachment and assign DS8000 volumes to VIOS Example 4 2 Creating a host attachment and assigning DS8000 volumes to VIOS dscli gt mkvolgrp hosttype pSeries vio_1 Residency CMUCO0030I mkvolgrp Volume group V15 successfully created dscli gt mkhostconnect wwname 10000000C9998F6E hosttype pSeries portgrp 60 p770_viol_0 CMUCO0012I mkhostconnect Host connection 0019 successfully created dscli gt mkhostconnect wwname 10000000C9998F6F hosttype pSeries portgrp 60 p770_viol_1 CMUCO0012I mkhostconnect Host connection 001A successfully created dscli gt mkfbvol cap 50 extpool p8 name p770_viol volgrp v15 9000 9002 CMUCO0025I mkfbvol FB volume 9000 successfully created CMUCO0025I mkfbvol FB volume 9001 successfully created CMUCO0025I mkfbvol FB volume 9002 successfully created dscli gt managehostconnect volgrp v15 60 CMUCOO016I managehostconnect Port group number 60 successfully modified To quickly check which volumes are assigned to the VIOS on a DS8000 storage system use the DS8000 Command Line Interface DS CLI with the 1shostconnect and 1sfbvol commands as shown in Example 4 3 The Ishostconnect login command even shows the DS8000 I O ports where the Fibre Channel adapters of the VIOS are logged in Example 4 3 Checking DS8000 volume assignment to VIOS using DSCLI dscli gt Ishostconnect portgrp 60 Name ID WWPN HostType Profile portgrp volgrpID ESSIOport p770
300. mple 6 51 Example 6 51 Mounting the source volume x36501ab9 mnt mount dev mapper 3600507630af fc29f0000000000004791 mnt itso_4791 x36501ab9 mnt mount dev mapper 3600507630af fc29f0000000000004791 on mnt itso_ 4791 type ext3 rw 2 Verify that the data on the source volume is consistent for example by running the sync command 3 Create the FlashCopy on the DS8000 and map the target volume to the Linux host In this example the FlashCopy source has the volume ID 4791 and the target volume has ID 47b1 Chapter 6 Linux considerations 111 4 Initiate a device scan on the Linux host For more information see Adding and removing DS8000 volumes dynamically on Linux on page 106 DM Multipath will automatically integrate the FlashCopy target as shown in Example 6 52 Example 6 52 Checking DM Multipath topology for target volume x36501ab9 mnt multipathd k show topology 3600507630af fc29f0000000000004791 dm 4 IBM 2107900 size 100G features 1 queue_if_no_path hwhandler 0 wp rw policy round robin 0 prio 1 status active 3 0 0 3 sde 8 64 active ready running 3 0 1 3 sdi 8 128 active ready running 4 0 1 3 sdq 65 0 active ready running 4 0 0 3 sdm 8 192 active ready running 3600507630af fc29f00000000000047b1 dm 5 IBM 2107900 size 100G features 1 queue_if_no_path hwhandler 0 wp rw policy round robin 0 prio 1 status active 3 0 1 4 sds 65 32 active ready running 3 0 0 4 sdr 65
301. mportant To avoid data integrity issues it is important that LVM configuration commands are not issued until step 5 is complete 5 Run the vgimportclone sh script against the target volumes providing a new volume group name as shown in Example 6 56 Example 6 56 Adjusting the target volumes LVM metadata x36501ab9 vgimportclone sh n vg_itso fc dev mapper 3600507630af fc29f00000000000046a0 dev mapper 3600507630af fc29f00000000000047 a0 WARNING Activation disabled No device mapper interaction will be attempted Physical volume tmp snap ptY8xojo vgimportl changed 1 physical volume changed O physical volumes not changed WARNING Activation disabled No device mapper interaction will be attempted Physical volume tmp snap ptY8xojo vgimportO changed 1 physical volume changed O physical volumes not changed WARNING Activation disabled No device mapper interaction will be attempted Volume group vg _ds8000 successfully changed Volume group vg_ds8000 successfully renamed to vg itso fc Reading all physical volumes This may take a while Found volume group vg_itso fc using metadata type lvm2 Found volume group vg ds8000 using metadata type lvm2 6 Activate the volume group on the target devices and mount the logical volume as shown in Example 6 57 Example 6 57 Activating volume group on the target device and mount the logical volume x36501ab9 vgchange a y vg _itso fc 1 logical volume s in volume grou
302. multiple hosts 12 5 2 Parallel access volumes PAV definition For EAV volumes do not consider using static Parallel Access Volumes PAV for future planning Instead for performance reasons use HyperPAV and you will be able to increase the amount of real device addresses within an LCU Two base concepts are basic in PAV functionality gt Base address The base device address in the conventional unit address of a logical volume There is only one base address associated with any volume gt Alias address An alias device address is mapped to a base address I O operations to an alias run against the associate s base address storage space There is no physical space associated with an alias address you can define more than one alias per base 262 IBM System Storage DS8000 Host Attachment and Interoperability Different usages of these aliases are implemented in different PAV typologies gt Static PAV means that the hosts can use only the aliases that were defined for that base address so if you need more parallels accesses to some devices a new HCD should be done to implements your new needs gt Dynamic PAV using Work Load Manager WLM the aliases are dynamically moved between different base address as requested by host jobs WLM also keeps track of the devices utilized by the different workloads accumulates this information over time and broadcasts it to the other systems in the same sysplex If WLM determines that any
303. n flight I Os for each file system At the LVM device driver layer hdisk buffers limit the number of in flight I Os At the SDD layer I Os are queued if the data path optimizer dpo device s attribute qdepth_enable is set to yes which it is by default The data path optimizer is a pseudo device which is the pseudo parent of the vpaths that allows you to control I O flow to SDD path devices Example 5 12 shows how to list the parameters for the dpo devices s attribute Example 5 12 Listing the qdepth_enable parameter for the dpo device s attribute lsattr El dpo 2062_max_luns 512 Maximum LUNS allowed for 2062 False 2105_max_luns 1200 Maximum LUNS allowed for 2105 True 2145_max_luns 512 Maximum LUNS allowed for 2145 False persistent_resv yes Subsystem Supports Persistent Reserve Command False qdepth_enable yes Queue Depth Control True Chapter 5 AIX considerations 69 With certain database applications such as an application running with an IBM DB2 database IBM Lotus Notes IBM Informix database the software might generate many threads which can send heavy I O to a relatively small number of devices Executing queue depth logic to control I O flow can cause performance degradation or even a system hang If you are executing queue depth logic use the qdepth_enable attribute to disable this queue depth logic on I O flow control This removes the limit on the amount of I O sent to vpath devices Some releases of SDD do no
304. n order to create these hostconnections identify the WWPNs of the servers HBAs One popular method of locating the WWPNs is to scan the var adm messages file Often the WWPN only shows up in the file after a restart Also the string to search for depends on the type of HBA that you have Specific details are available in the IBM System Storage DS8000 Host Systems Attachment Guide SC26 7917 In many cases you can also use the prtconf command to list the WWPNs Example 9 1 Example 9 1 Listing the WWPNs prtconf vp grep port wwn port wwn 21000003 ba43fdcl port wwn 210000e0 8b099408 port wwn 210000e0 8b0995f3 port wwn 210000e0 8b096cf6 port wwn 210000e0 8b098f08 160 IBM System Storage DS8000 Host Attachment and Interoperability With Solaris 10 the fcinfo command provides the ability to find the WWPN immediately as shown in Example 9 2 Example 9 2 Using fcinfo to find the WWPNs fcinfo hba port HBA Port WWN 10000000c95f3592 OS Device Name dev cfg c4 Manufacturer Emulex Model LP10000 Firmware Version 1 92a1 FCode BIOS Version none Type L port State offline Supported Speeds 1Gb 2Gb Current Speed not established Node WWN 20000000c95f3592 HBA Port WWN 10000000c95f35dc OS Device Name dev cfg c3 Manufacturer Emulex Model LP10000 Firmware Version 1 92a1 FCode BIOS Version none Type unknown State offline Supported Speeds 1Gb 2Gb Current Speed not established Node WWN 20000000c95f35dc 9 3
305. n softcopy only gt gt gt gt DS8000 Performance Monitoring and Tuning SG24 7146 DS8000 Thin Provisioning REDP 4554 IBM PowerVM Virtualization Managing and Monitoring SG24 7590 IBM System z Connectivity Handbook SG24 5444 IBM System Storage DS8000 LDAP Authentication REDP 4505 IBM System Storage DS8700 Easy Tier REDP 4667 Migrating to IBM System Storage DS8000 SG24 7432 PowerVM and SAN Copy Services REDP 4610 PowerVM Virtualization on IBM System p Introduction and Configuration Fourth Edition SG24 7940 Multiple Subchannel Sets An Implementation View REDP 4387 You can search for view or download Redbooks publications Redpaper publications Technotes draft publications and Additional materials as well as order hardcopy Redbooks publications at this website ibm com redbooks Other publications These publications are also relevant as further information sources gt gt gt gt Device Support Facilities User s Guide and Reference GC35 0033 IBM System Storage DS8000 Command Line Interface User s Guide SC26 7916 IBM System Storage DS8000 Host Systems Attachment Guide SC26 7917 IBM System Storage DS8000 Introduction and Planning Guide GC35 0515 IBM System Storage DS Open Application Programming Interface Reference GC35 0516 IBM System Storage Multipath Subsystem Device Driver User s Guide GC52 1309 Copyright IBM Corp 2011 2013 All rights reserved 285 Online
306. n this case you do not need to use the Diskpart tool Use the Windows Disk Management functions to allocate the new space Expansion works and is not dependent on the volume type for example simple spanned mirrored and so on on the disk Dynamic disks can be expanded without stopping the I O in most cases Important Always back up your data before upgrading your basic disk to dynamic disk or in reverse order as this operation is disruptive to the data due to the separate position of the logical block address LBA in the disks 3 8 SAN boot support When booting from the Fibre Channel storage systems special restrictions apply gt With Windows Server 2003 and 2008 MSCS uses target resets See the Microsoft technical article Microsoft Windows Clustering Storage Area Networks at the following website http support microsoft com kb 818668 gt Windows Server 2003 and 2008 enable boot disk and the cluster server disks to be hosted on the same bus However you need to use Storport Miniport HBA drivers for this functionality to work The boot disk and cluster server disks hosted on the same bus is not a supported configuration in combination with drivers of other types for example SCSlport Miniport or full port drivers gt If you reboot a system with adapters while the primary path is in a failed state manually disable the BIOS on the first adapter and manually enable the BIOS on the second adapter You cannot enable
307. nd as shown in Example 3 9 Example 3 9 Expanding a volume dscli gt chfbvol cap 18 4703 CMUC00332W chfbvol Some host operating systems do not support changing the volume size Are you sure that you want to resize the volume y n y CMUCO0026I chfbvol FB volume 4703 successfully modified The new capacity must be larger than the previous one you cannot shrink the volume To ensure that the volume was expanded use the 1sfbvol command as shown in Example 3 10 In this example you can see that the volume with the ID of 4703 was expanded to 18 GB in capacity Example 3 10 Isfbvol after expansion dscli gt Isfbvol 4703 Name ID accstate datastate configstate deviceMTM datatype extpool cap 2 30B cap 10 9B cap blocks ITSO_x346_3_4703 4703 Online Normal Normal 2107 900 FB 512 P53 18 0 37748736 Chapter 3 Windows considerations 29 In Disk Management perform a rescan for the disks to see the new capacity for disk1 Figure 3 9 shows the disks available after a rescan is performed File Action view Window Help e System Tools J C Partition Basic NTFS Healthy Boot 68 36 GB 62 00 GB E Event Viewer J E Partition Basic NTFS Healthy System 33 90 GB 33 83 GB Shared Folders 9D58000_ExP Z Partition Basic NTFS Healthy 12 00GB 11 93 GB Local Users and Groups Performance Logs and Alert a Device Manager Sy Storage Removable Storage Disk Defragmenter Disk Management
308. nd supports all block devices that can have more than one path IBM has also adopted a strategy to support DM Multipath wherever possible Tip The SDD the IBM proprietary multipathing solution is no longer updated for Linux Adding and removing volumes online With the new hotplug and udev subsystems it is now possible to easily add and remove disks from Linux SAN attached volumes are usually not detected automatically because adding a volume to a DS8000 volume group does not create a hotplug trigger event such as inserting a USB storage device SAN attached volumes are discovered during the user initiated device scans and then automatically integrated into the system including multipathing To remove a disk device before you physically detach it make sure it is not in use and then remove it logically from the system Dynamic LUN resizing Recently improvements were introduced to the SCSI layer and DM Multipath that allow resizing of SAN attached volumes when they are in use Capabilities are still limited to certain cases 6 2 Attaching to a basic host In this section information is provided for making DS8000 volumes available to your Linux host Additionally separate methods of attaching storage for separate hardware architectures are explained in addition to configuring the Fibre Channel HBA driver setting up multipathing and any required special settings 6 2 1 Platform specific information The most popular hardware pl
309. nd write commands indicate a transfer length and the control unit returns a residual count on the last command in the command chain zero if all data set gt Original ZHPF delivery z10 did not store any data in memory on reads if residual transfer count was non zero gt Newer ZHPF channels indicate they have ZHPF incorrect length support in I O port login PERLI gt If supported control unit presents non zero residual counts without presenting unit check gt Needed for QSAM and BSAM gt Supported on z OS V1R11 V1R12 V1R13 and above 270 IBM System Storage DS8000 Host Attachment and Interoperability ZHPF bi directional data transfer and TCA extension Original ZHPF supported commands transferred limited to about 240 bytes in a command information unit Now ZHPF support for TCA extension allows more commands per I O operation gt Allows up to about 1000 bytes for commands in beginning of data transfer gt Allows for long command chains to be issued in a single command gt ZHPF channel supports bi directional data transfer to allow read data after writing TCA extension See Figure 12 4 Channel Device Write Type Commands Command ui TCA Extension Write Data l Read Type Commands Command Ul TCA Extension Read Data Figure 12 4 zHPF bi directional data transfer example gt Needed by DB2 which reads up to 32 pages per I O operation gt Supported on z
310. ndow 140 IBM System Storage DS8000 Host Attachment and Interoperability Using NPIV with RDMs More information about NPIV can be found in the following VMware publications http www vmware com files pdf techpaper vsp 4 vsp4 41 npivconfig pdf Attention NPIV can only be used with RDMs For more granular QoS measurement and monitoring a virtual WWN can be created for a VM All I O from this VM will be done using this virtual WWN To use NPIV with a virtual machine perform the following steps Important The virtual machine needs to be powered off for these steps 1 Create an RDM as described in Assigning an RDM to a VM on page 139 2 In the Virtual Machine Properties window of the VM select the Options tab and click Fibre Channel NPIV as shown in Figure 7 38 Remove the checkmark from Temporarily Disable select Generate new WWNs and choose the amounts that you want Resources Profiles vServices Virtual Machine Version Settings Summary m Fibre Channel Virtual WWNs General Options SUSE VM Virtual machines running on hosts with Fibre Channel hardware von Oo Sate SRG eS eee VMware Tools Shut Down petra oa lis id y a Power Management Stan s dby 7 Temporarily Disable NPIV for this virtual machine Advanced General Normal No WWNs are currently assigned CPUID Mask Expose Nx flagto Boot Options Normal Boot C Leave unchanged CFibre Channel NPIV Generating Generate new W
311. ndow Help 18 xl e m e mage Volume Layout Type File System status Capacity Free Space Free GJ C Partition Basic NTFS Healthy Boot 68 36GB 62 18GB 90 CI E Partition Basic NTFS Healthy System 33 90 GB 33 83GB 99 Computer Management Local System Tools e Event Viewer Shared Folders x Local Users and Groups a Performance Lags and Alert a gt Device Manager ZPDisk 2 Sj Storage Dynamic Removable Storage 12 00 GB 12 00 GB Disk Defragmenter Online Unallocated Disk Management f BR Disk 3 n E Services and Applications Dynamic 12 00 GB 12 00 GB Online Unallocated SSDisk 4 A Dynamic 12 00 GB 12 00 GB Online Unallocated eee 4 gt BB Unallocated i Primary partition Figure 3 5 Initializing disks in the Disk Management view Tip Newly assigned disks will be discovered If they are not discovered go to the Disk Manager window and rescan disks or go to the Device Manager window and scan for hardware changes 24 IBM System Storage DS8000 Host Attachment and Interoperability 3 6 2 SDDDSM datapath query Example 3 3 shows the output of the datapath query device command Example 3 3 datapath query device with SDDDSM C Program Files IBM SDDDSM gt datapath query device Total Devices 3 DEV 0 DEVICE NAME Disk2
312. ned volumes The file system must be created directly on the DM Multipath device gt Only the modern file systems can be resized when they are mounted The ext2 file system cannot 6 4 Using FlashCopy and remote replication targets DS8000 FlashCopy and remote replication solutions such as IBM Metro Mirror or IBM Global Mirror create bitwise identical copies of the source volumes The target volume has a separate unique identifier which is generated from the DS8000 WWNN and volume ID Any metadata that is stored on the target volume such as the file system identifier or LVM signature is identical to that of the source volume This can lead to confusion and data integrity problems if you plan to use the target volume on the same Linux system as the source volume Information presented in this section describe methods that avoid integrity issues when you use the target volume on the same Linux system as the source volume 6 4 1 Using a file system residing on a DS8000 volume The copy of a file system created directly on a SCSI disk device or single path or a DM Multipath device without an additional virtualization layer such as RAID or LVM can be used on the same host as the source without modification To use this copy of a file system complete the following steps 1 Mount the original file system by using a device node that is bound to the volume s unique identifier and not to any metadata that is stored on the device itself Exa
313. ned with QUERY PAV rdev rdev where xx is the channel path identifier CHPID whose device addresses is displayed showing the addresses and any aliases and rdev is the real device address PAV minidisk support with SPE Starting with z VM V5 2 0 with APAR VM63952 z VM supports PAV minidisks With SPE z VM provides the following advantages gt gt Support of PAV minidisks Workload balancing for guests that do not exploit PAV such as Conversational Monitor System A real I O dispatcher queues minidisk I O across system attached alias volumes Minidisks that can be linked for guests that exploit PAV for example z OS and Linux The PAVALIAS parameter of the DASDOPT and MINIOPT user directory statements or the CP DEFINE ALIAS command which creates alias minidisks both fullpack and non fullpack for using guests Dynamic alias to base reassociation is supported for guests that exploit PAV for dedicated volumes and for minidisks under restricted conditions HyperPAV support Starting with z VM Version 5 4 z VM supports HyperPAV for dedicated DASD and minidisks 12 6 4 Missing interrupt handler The z VM operating system sets its MIH value to 1 25 multiplied by what the hardware reports This way with the DS8000 setting the MIH value to 30 seconds z VM is set to a MIH value of approximately 37 5 seconds It enables the guest to receive the MIH 30 seconds before z VM does 12 7 VSE ESA and 2 VSE considerations The following
314. nel Sets An Implementation View REDP 4387 IBM System Storage DS8000 Host Attachment and Interoperability New Channel Subsystems CSS were also implemented on z10 z196 and z114 See the related installation and planning guide to verify total CSS availability on your host For more details about the IBM zEnterprise System see the Redbooks publication BM zEnterprise System Technical Introduction SG24 7832 Important gt The logical configuration on the host HCD must match with the device logical configuration in order to be able to use the devices gt A layout of correspondence within UCBs host and device LUNs also needs to be kept updated to expedite each analysis to be done on your environment Read control unit and device recognition for DS8000 The host system informs the attached DS8000 of its capabilities so that it supports native DS8000 control unit and devices The DS8000 then only returns information that is supported by the attached host system using the self description data such as read data characteristics sense ID and read configuration data The following commands are able to recognize device type 2107 in their output gt DEVSERV QDASD and PATHS command responses gt IDCAMS LISTDATA COUNTS DSTATUS STATUS and IDCAMS SETCACHE 12 3 3 zDAC z OS FICON Discovery and Auto Configuration feature The z OS FICON Discovery and Auto Configuration ZDAC feature which is deployed by the new z Enterprise
315. nel disk controller PCI x 5749 4 GB Fibre Channel disk controller PCI x 5774 4 GB Fibre Channel disk controller PCI Express 5735 8 GB Fibre Channel disk controller PCI Express 5273 PCle LP 8 Gb 2 Port Fibre Channel adapter 5276 PCle LP 4 Gb 2 Port Fibre Channel adapter vvvvvvy Adapters with feature numbers 2787 and 5760 are 1 port input output processor IOP based adapters that can address up to 32 logical volumes The other adapters are 2 port IOP less adapters that can address up to 64 volumes per port The Low Profile adapters feature number 5273 and 5276 are supported in POWER models that support Low Profile PCI slots For more information see PCI adapter placement rules and slot priorities in the IBM Power Systems Hardware Information Center as listed in 11 1 4 Useful websites on page 204 Chapter 11 IBMiconsiderations 205 Tip You can assign up to 32 LUNs for two or more IOP based adapters in multipath or up to 64 LUNs for two or more ports in IOP less adapters The ports in multipath must belong to separate IOP less adapters IBM i V6R1 and later in combination with IOP less adapters provides SCSI command tag queuing support on DS8000 systems with the supported queue depth being six input and output operations 11 2 2 Attachment with VIOS NPIV Connecting DS8000 series to an IBM i with VIOS NPIV in POWER servers requires one of the following adapters in VIOS gt 5735 8 GB Fibre Channel disk
316. ng system environment for example Solaris installed on an internal disk device Example 9 20 shows the boot device and the mount points of such an environment Example 9 20 Current setup root prime 1ab 04 export home test eeprom grep boot device boot device pci 83 4000 FJSV ulsa 2 1 disk 0 0 a disk root prime 1lab 04 format Searching for disks done AVAILABLE DISK SELECTIONS 0 3 2 cOt0d0 lt FUJITSU MAW3073NC 3701 cyl 24345 alt 2 hd 8 sec 737 gt pci 83 4000 FJSV ulsa 2 1 sd 0 0 clt0d0 lt FUJITSU MAW3073NC 3701 cyl 24345 alt 2 hd 8 sec 737 gt pci 83 4000 FISV ulsa 2 sd 0 0 c3t500507630A03029Fd0 lt IBM 2107900 268 cyl 4367 alt 2 hd 30 sec 64 gt pci 80 2000 fibre channel 2 fp 0 0 ssd w500507630a03029F 0 3t500507630A03029Fd1 lt IBM 2107900 268 cyl 8736 alt 2 hd 30 sec 64 gt pci 80 2000 fibre channel 2 fp 0 0 ssd w500507630a03029F 1 3t500507630A03029Fd2 lt IBM 2107900 268 cyl 17918 alt 2 hd 64 sec 256 gt pci 80 2000 fibre channel 2 fp 0 0 ssd w500507630a03029F 2 C4t500507630A00029Fd0 lt IBM 2107900 268 cyl 4367 alt 2 hd 30 sec 64 gt pci 80 2000 fibre channel 1 fp 0 0 ssd w500507630a00029F 0 C4t500507630A00029Fd1 lt IBM 2107900 268 cyl 8736 alt 2 hd 30 sec 64 gt pci 80 2000 fibre channel 1 fp 0 0 ssd w500507630a00029F 1 C4t500507630A00029Fd2 lt IBM 2107900 268 cyl 17918 alt 2 hd 64 sec 256 gt pci 80 2000 fibre channel 1 fp 0 0 ssd w500507630a00029F
317. ng with DS8000 licensed machine code 5 3 xx xx it is possible to stripe the extents of a DS8000 logical volume across multiple RAID arrays DS8000 storage pool striping can improve throughput for certain workloads It is performed on a 1 GB granularity so it will generally benefit random workloads more than consecutive workloads If you are already using a host based striping method for example LVM striping or DB2 database container striping you do not need to use storage pool striping but it is possible If you use a host based striping method and a storage pool striping method then combine the wide stripes on DS8000 with small granularity stripes on the host The preferred size for these is usually between 8 and 64 MB If large stripes on both DS8000 and attached host interfere with each other I O performance might be affected 5 6 Using AIX access methods for I O AIX provides several modes to access data in a file system It can be important for performance to select the right access method Use the information provided in this section to help determine your access method 5 6 1 Synchronous I O Synchronous I O occurs while you wait An application s processing cannot continue until the I O operation is complete It is a more secure and traditional way of handling data It helps ensure consistency at all times but can affect performance It also does not allow the operating system to take full advantage of functions of modern storage d
318. nline_Storage_Rec onfiguration_Guide index html gt Red Hat 6 http docs redhat com docs en US Red_Hat_Enterprise_Linux 6 pdf Storage Adminis tration_Guide Red_Hat_Enterprise_Linux 6 Storage_Administration_Guide en US pdf DM Multipath Configuration and Administration The Red Hat document DM Multipath Configuration and Administration is also part of the information provided for Red Hat Enterprise Linux 5 In this document you can find general information for Red Hat Enterprise Linux 5 in addition to other Linux operating systems for device mapper multipathing DM Multipath such as the following topics gt How DM Multipath works gt How to set up and configure DM Multipath within Red Hat Enterprise Linux 5 gt Troubleshooting DM Multipath This documentation is available at the following website http docs redhat com docs en US Red_Hat_Enterprise_Linux 5 htm1 DM_Multipath ind ex html For Red Hat Enterprise 6 DM Multipath documentation see the following website http docs redhat com docs en US Red_Hat_Enterprise_Linux 6 pdf DM_Multipath Red_ Hat_Enterprise_Linux 6 DM Multipath en US pdf SLES 11 SP2 Storage Administration Guide The SLES 11 SP2 Storage Administration Guide is part of the documentation for Novell SUSE Linux Enterprise Server 11 Service Pack 2 Although written specifically for SUSE Linux Enterprise Server this guide contains useful storage related information for Linux such as the following topics
319. nnect to the IBM i in multipath chdev dev hdiskX attr reserve_policy no_reserve 2 Set the attributes of Fibre Channel adapters in the VIOS to fc_err_recov fast_fail and dyntrk yes When the attributes are set to these values the error handling in Fibre Channel adapter enables faster transfer to the alternate paths in case of problems with one Fibre Channel path To help make multipath within one VIOS work more efficient specify these values by entering the following command chdev dev fscsix attr fc_err_recov fast_fail dyntrk yes perm 3 Enter the following command for each hdisk hdiskX to get more bandwidth by using multiple paths chdev dev hdiskX perm attr algorithm round_robin Ensure that the queue depth in VIOS and the queue depth in the VIOS matches the IBM i queue depth of 32 by using the following commands 1 Enter the following VIOS command to check the queue depth on physical disks lsdev dev hdiskxx attr queue_depth 2 Set the queue depth to 32 if needed by using the following command chdev dev hdiskxx attr queue _depth 32 VIOS NPIV Set the attribute of Fibre Channel adapters in VIOS to fc err recov fast and dyntrk yes perm as is explained in VIOS VSCSI on page 224 11 6 8 Adding volumes to the System i configuration After the logical volumes are created and assigned to the host they appear as non configured units to IBM i When assigning a large quantity of LUNs it can take time for all of
320. ns config config log lockrgn Multipathi numpaths priv 000048 000239 000192 copy 01 offset 000000 enabled priv 000256 048207 047952 copy 01 offset 000192 enabled priv 048208 055503 007296 copy 01 offset 000000 enabled priv 055504 055647 000144 part 00 offset 000000 ng information 2 c2t50050763030CC1A5d0s2 state enabled c3t50050763031CC1A5d0s2 state enabled 170 IBM System Storage DS8000 Host Attachment and Interoperability After the volume expansion the disk size is 37677568 When multiplied by 512 bytes it equals 18 GB Tip You need at least two disks in the diskgroup where you want to resize a disk otherwise the vxdisk resize command will fail To expand the logical volume and the file system in VxVM first you need to determine the maximum size that you can expand to then you can expand the logical volume and the file system To find the maximum size that you can expand to issue the vxvoladm command as shown in Example 9 18 Example 9 18 VxVM logical volume expansion bash 3 00 vxvoladm g 20781 dg maxgrow 20781 vol Volume can be extended to 37677056 17 97g bash 3 00 vxvoladm g 20781_dg growto 20781 vol 37677056 bash 3 00 opt VRTS bin fsadm b 17g 20781 UX vxfs fsadm INFO V 3 25942 dev vx rdsk 20781_dg 20781_vol size increased from 23068672 sectors to 35651584 sectors After the file system expansion the df command shows a size of 17825792 KB equal to approximately 17 GB on file syste
321. ns within HP UX there is a maximum number that can be assigned when configuring LUNs ID Numbers greater than x 3FFF Decimal 16383 are not supported It creates a limitation of 16 384 LUNs that can be enumerated on the host See the IBM Information Center section on HP UX host attachment for more information Important LUN IDs cannot exceed x 3FFF http publib boulder ibm com infocenter dsichelp ds8000ic index jsp topic com ibm storage ssic help doc f2c_attchnghpux_1t xvv html 10 2 Available resources For the latest available supported HP UX configuration and required software patches see the IBM SSIC at the following website http www 03 ibm com systems support storage ssic interoperability wss For information about how to prepare the host to attach the DS8000 see the DS8000 Information Center at the following website http publib boulder ibm com infocenter dsichelp ds8000ic index jsp In the Information Center select Configuring Attaching Hosts gt HP UX host attachment To get the latest version of the DS CLI to install on your host use the version delivered with the DS8000 Microcode Bundle Alternatively download the latest available International Organization for Standardization ISO image for the DS CLI client download at this website ftp ftp software ibm com storage ds8000 updates DS8K_Customer_Download Files CLI Chapter 10 HP UX considerations 183 10 3 Identifying available HBAs
322. nto significantly fewer sequences The effort required to convert individual commands into FICON format is removed as multiple System z I O commands are packaged together and passed directly over the fiber optic link The ZHPF system provides an enhanced FICON protocol and system I O architecture that results in improvements for small block transfers with a track or less to disk using the device independent random access method In situations where it is the exclusive access in use it can help improve FICON I O throughput on a single DS8000 port by 100 Realistic workloads with a mix of data set transfer sizes can see 30 70 of FICON I Os utilizing ZHPF with potential results of up to a 10 30 channel utilization savings Although clients can see I Os complete faster as the result of implementing ZHPF the real benefit is expected to be obtained by using fewer channels to support existing disk volumes or increasing the number of disk volumes supported by existing channels Additionally the changes in architectures offer end to end system enhancements to improve reliability availability and serviceability The ZHPF system uses multitrack operations to allow reading or writing of more than a track s worth of data by a single transport mode operation DB2 VSAM zFS the hierarchical file system partitioned data set extended PDSE striped extended format data sets and other applications that use large data transfers are expected to bene
323. nts welcome 00000 tnt tne xiii Stay connected to IBM Redbooks 0 00 e eee eee eee xiii Chapter 1 General considerations 0000 c cette eee 1 1 1 DS8000 general topics 1 0 0 teens 2 1 2 Client architecture overview 0 0 000 tenes 3 1 3 SAN considerations 0 00 teens 3 TA Storage MOIS e aare Seale a hale BAT Rk BEE PE BN IE week 3 Chapter 2 Open systems considerations 000s 5 2 1 Configuration resources sssusa auauua 6 2 1 1 IBM TESOUCOS res eiiiai a e a a aa a a R a Dae 6 2 1 2 HBA vendor resources 1 2 00 ce teas 7 2 2 Using the DS8000 as a boot device 0 8 2 2 1 Configuring the QLogic BIOS to boot from a DS8000 volume 9 252 2 N Xt SIOPS onire a iia eect et Pee eR ane ae reed vee ae eae Bt 13 2 3 Additional supported configurations sasssa cee 13 2 4 Multipathing support for Subsystem Device Drivers 00 000 eee 14 2 4 1 Subsystem Device Driver 0 000 cette 14 2 4 2 Other multipathing solutions 0 000 c eee 15 Chapter 3 Windows considerations 00 000 e cee eee es 17 3 1 Attaching HBAs in Windows 0 tees 18 3 1 1 HBA and operating system settings 0 00 0 eee ee 18 3 1 2 SDD versus SDDDSM Multipath Drivers 00 00 18 3 2 Installing SDD in Windows 0 0c tenes 18 3 2 1 SDD datapath query 1 0 0 0 tees 20 3 2 2 Map
324. o Technology home page is found at the following website http www attotech com You can find the Atto Technology support page at the following website http www attotech com solutions ibm html You must register with Atto Technology to download drivers and utilities for HBAs Platform and operating system vendor pages The platform and operating system vendors also provide support information for their clients See their information for general guidance about connecting their systems to SAN attached storage If you cannot find information that can help you with third party vendors check with your IBM representative about interoperability and support from IBM for these products It is beyond the scope of this book to list all of the vendor websites 2 2 Using the DS8000 as a boot device For most of the supported platforms and operating systems you can use the DS8000 as a boot device The IBM SSIC provides detailed information about boot support for specific operating systems For open systems the first step is to configure the HBA to load a BIOS extension It provides the basic input and output capabilities for a SAN attached disk assuming that the host is equipped with QLogic HBAs Tip For IBM branded QLogic HBAs the BIOS setup utility is called FAST Util 8 IBM System Storage DS8000 Host Attachment and Interoperability 2 2 1 Configuring the QLogic BIOS to boot from a DS8000 volume To enable the QLogic BIOS to boot from a
325. o two or more virtual adapters each virtual adapter assigned to a separate VIOS This way resiliency for the boot disk is provided even if one VIOS fails To achieve such multipathing you need VIOS level 2 1 2 or later 11 7 2 Tagging the load source LUN To enable IBM i to boot from an external load source the adapter that is to be used for this external LUN must be tagged as the load source I O device For IBM i in a partition of a POWER server or System p server the external load source is tagged in the HMC in the profile of the IBM i LPAR With natively connected DS8000 tag the physical adapter connecting the load source LUN With DS8000 connected through VIOS tag the virtual adapter to the boot LUN that is assigned or mapped Chapter 11 IBM i considerations 235 To tag an adapter as the load source I O device in HMC complete the following steps 1 In HMC expand the pop up menu at the IBM i partition and select Configuration gt Manage profiles Figure 11 27 Systems Management gt Servers gt p6 570 2 06C6DE1 e lel ee srt gt iwr SS a Processing Units Memory G8 By aTs_mS0_GERO_DUALVIOS 1_1P38 Oo 15 Running O Ei ats_ceari_mso_virt p34 11 Running O El ats_tpar12_mso_virt_1p3s 12 Running M E ats tears _mTso_sana_ p42 Properties Change Defautt Profile Operations Configuration Hardware Information Dynamic Logical Partitioning Servicea
326. olume IDs IBM System Storage DS8000 Host Attachment and Interoperability The installed multipathing file sets on the VIOS can be checked using the 1ssw command as shown in Example 4 4 Example 4 4 Checking for the installed SDD or SDDPCM file set on the Virtual I O Server Issw grep iE sdd fcp disk ibm FCP Disk devices fcp disk ibm mpio rte 1 0 0 21 C F IBM MPIO FCP Disk Device devices sddpcm 61 rte 2 6 0 0 C F IBM SDD PCM for AIX V61 There is no efix data on this system For a redundant VIOS setup Figure 4 3 with two VIOS partitions providing the same DS8000 LUNs to a client partition using multipathing only SDDPCM is supported on the VIOS VIO Client Virtual SCSI Disk active path ia _ Passive path DS8000 Figure 4 3 Dual VIOS setup with DS8000 backing devices and multipath support in client partition For more information see Chapter 4 section 4 14 1 Supported virtual SCSI configurations in the section Supported scenarios using multiple Virtual I O Servers in the PowerVM Virtualization on IBM System p Introduction and Configuration Fourth Edition SG24 7940 IBM i multipath is automatically started on the DS8000 LUNs that are connected through two VIOS partitions in virtual SCSI There is no need to install any particular driver for multipath in IBM i For proper multipath support of virtual SCSI devices in the client partition with a dual VIOS setup the DS8000 LUN m
327. ompt Chapter 6 Linux considerations 105 Disabling QLogic failover The QLogic HBA kernel modules have limited built in multipathing capabilities Because multipathing is managed by DM Multipath ensure that the Qlogic failover support is disabled Use the modinfo qla2xxx command to verify as shown in Example 6 35 Example 6 35 Verify QLogic failover is disabled x36501ab9 modinfo qla2xxx grep version version 8 03 01 04 05 05 k srcversion A2023F2884100228981F34F If the version string ends with fo the failover capabilities are turned on and must be disabled To do this add a line to the etc modprobe conf file of your Linux system as shown in Example 6 36 Example 6 36 Disabling QLogic failover x36501ab9 cat etc modprobe conf options qla2xxx ql2xfailover 0 After modifying this file run the depmod a command to refresh the kernel driver dependencies Then reload the qla2xxx module to make the change effective If you have included the qla2xxx module in the InitRAMFS you must create a new qla2xxx module 6 2 10 Understanding non disruptive actions on attached hosts This section describes some non disruptive actions that can be taken on attached hosts Adding and removing DS8000 volumes dynamically on Linux Adding and removing DS8000 volumes in Linux on System z Adding new DS8000 host ports to Linux on System z Resizing DS8000 volumes dynamically Using FlashCopy and remote replication targets YYYY Y Adding
328. on a given AIX server not both Also notice that the separate SDD or SDDPCM file sets are dependent on the AIX release See the SDD or SDDPCM support matrix at the following website for further information and file set downloads http www ibm com support docview wss rs 540 amp context ST52G7 amp dc DA4008ui d ssg1S700 1350 Also see the readme file of your SDD or SDDPCM version for latest updates and prerequisites New SDD or SDDPCM releases typically require a specific minimum version of the host attachment file set For supported SDD and SDDPCM releases with IBM System Storage DS8000 series environments on various AIX releases see the IBM SSIC website at this website http www 03 ibm com systems support storage ssic interoperability wss For the management of MPIO devices in AIX 6 1 see the following online documentation http publib boulder ibm com infocenter aix v6rl index jsp topic com ibm aix bas eadmn doc baseadmndita dm_mpio htm 5 4 1 SDD for AIX SDD to SDDPCM migration AIX Version 7 1 does not support the IBM Subsystem Device Driver SDD for IBM TotalStorage IBM Enterprise Storage Server the IBM TotalStorage DS family and the IBM System Storage SAN Volume Controller If you are using SDD you must transition to Subsystem Device Driver Path Control Module SDDPCM or AIX Path Control Module PCM for the multipath support on AIX for IBM SAN storage SDD to SDDPCM migration scripts are available to help you with the
329. operating system can work with To create the logical volumes create file systems and mount them by completing the following steps a Create a logical volume which creates a 50 MB logical volume named 1vol1 lvcreate L 50 dev vg08 b Create a file system newfs F vxfs dev vg08 rlvol1 188 IBM System Storage DS8000 Host Attachment and Interoperability c Mount the file system mount dev vg01 lvol1 mountl This step assumes that the directory where the file system will be mounted already exists 3 Create the volume group To use the new agile addressing with a volume group specify the persistent DSFs in the vgcreate command as shown in Example 10 8 Example 10 8 Volume group creation with Persistent DSFs vgcreate A y x y 1 255 p 100 s 16 dev vgagile dev disk disk12 dev disk disk13 dev disk disk14 Volume group dev vgagile has been successfully created Volume Group configuration for dev vgagile has been saved in etc lvmconf vgagile conf vgdisplay v vgagile Volume groups VG Name VG Write Access VG Status Max LV Cur LV Open LV Max PV Cur PV Act PV Max PE per PV VGDA PE Size MBytes Total PE Alloc PE Free PE Total PVG Total Spare PVs Total Spare PVs in use Physical volumes PV Name PV Status Total PE Free PE Autoswitch PV Name PV Status Total PE Free PE Autoswitch PV Name PV Status Total PE Free PE Autoswitch dev vgagile
330. operating systems You can download it from this website ftp ftp software ibm com storage subsystem UG 1 8 3 0 Important IBM does not provide an SDD for each operating system described in this book For instance an SDD for HP UX 1 1iv3 is not available 14 IBM System Storage DS8000 Host Attachment and Interoperability 2 4 2 Other multipathing solutions Certain operating systems come with native multipathing software such as the following examples gt StorEdge Traffic Manager for Oracle Solaris This software is for Solaris 10 only If you use an older Solaris version you must explicitly download this software gt HP UX in version 11iv3 While HP still provides PVLinks for HP UX 1 iv it is generally not advisable to use PVLinks but rather the native multipathing that is provided with the operating system Native multipathing can help balance the I O load across multiple paths gt IBM AIX native multipathing I O support MPIO gt IBM i multipathing support In addition third party multipathing solutions are available such as Veritas DMP which is part of Veritas Volume Manager Most of these solutions are also supported for DS8000 attachment although the scope can vary Certain HBAs or operating system versions might have limitations For the latest information see the IBM SSIC at this website http www ibm com systems support storage config ssic Chapter 2 Open systems considerations 15 16 IBM Syst
331. or System z V5 1 Installation and Configuration Guide SC27 4053 gt IBM Tivoli Storage Productivity Center for Replication for System z V5 1 User s Guide C27 4054 12 9 1 Tivoli Storage Productivity Center for Replication for System z TPC R The TPC R for System z combines the former Two Site BC and Three Site BC Business Continuity offering into a single license It provides support for Metro Mirror and Global Mirror configurations as well as three site recovery management supporting IBM System Storage DS8000 Metro Global Mirror and Metro Global Mirror with HyperSwap It is designed to support fast failover and failback fast reestablishment of three site mirroring data currency at the remote site with minimal lag behind the local site and quick re synchronization of mirrored sites using incremental changes only Tivoli Storage Productivity Center for Replication for System z products support these features FlashCopy Basic HyperSwap Metro Mirror Metro Mirror with HyperSwap Metro Mirror with Open HyperSwap Global Mirror IBM System Storage DS8000 Metro Global Mirror IBM System Storage DS8000 Metro Global Mirror with HyperSwap vvvvvvvy New configuration support Tivoli Storage Productivity Center for Replication for System z provides support for IBM DS8000 V6 3 Extent Space Efficient volumes in all Remote Copy Services relationships Combined with DS8000 V6 3 Users can now use Extent Space Efficient ESE volumes in any
332. orted currently with host adapters QLA23xx 32 bit or 64 bit and LP9xxx 32 bit only 3 1 1 HBA and operating system settings Depending on the HBA type several HBA and driver settings might be required For a complete description of these settings see the IBM System Storage DS8000 Host Systems Attachment Guide SC26 7917 Although you can access the volumes by using other settings the values provided in this guide were tested Set the Time Out Value parameter associated with the host adapters to 60 seconds when attaching a storage unit to one of the following Microsoft systems gt Windows 2003 Server gt Windows 2008 Server This helps ensure optimum availability and recoverability The operating system uses the Time Out Value parameter to bind its recovery actions and responses to the disk subsystem The value is stored in the Windows registry in the HKEY_LOCAL_MACHINE SYSTEM CurrentControlSet Services Disk TimeOutValue path The value has the data type REG DWORD and must be set to 0x0000003c hexadecimal 60 decimal 3 1 2 SDD versus SDDDSM Multipath Drivers There are some big differences to point out between SDD and SDDDSM Multipath Drives If you are installing Windows Server 2003 it is highly advisable to move to SDDDSM because SDD is an older application version compatible with Windows 2000 Servers which is no longer supported by IBM If you install SDD in Windows Server 2003 you will have many multipathing limitations wh
333. otected versus unprotected volumes on page 221 On DS8000 models prior to DS8800 specify parameter eam rotateext to achieve storage pool striping for IBM i LUNs For the DS8800 rotate extents is the default extent allocation method For example you can use a DS CLI command to create an IBM i volume on a DS8800 mkfbvol extpool p6 os400 A04 name ITSO_i_ h 2000 2003 This command defines a protected volume of size 70 56 GB with default rotate extents method If the DS8000 is to be attached to IBM i with VIOS VSCSI create LUNs as for open systems Such LUNs can be of any size and you do not need to define them in one of the fixed sizes for IBM i However consider keeping the size of volumes within the advisable limits as is explained in 11 4 Sizing and numbering of LUNs on page 215 11 6 3 Protected versus unprotected volumes When defining IBM i logical volumes you must decide whether these are to be protected or unprotected It is simply a notification to IBM i and does not mean that the volume is protected or unprotected In fact all DS8700 LUNs are protected by either RAID 5 RAID 6 or RAID 10 Defining a volume as unprotected means that is available for IBM i to perform a mirroring of that volume to another of equal capacity either internal or external If you do not intend to use IBM i host based mirroring define your logical volumes as protected Chapter 11 IBM i considerations 221 11 6 4 Changing LUN protection
334. ou can use either direct or switched attachment to attach a storage unit to a System z host system that runs Novell SUSE Linux Enterprise Server10 SP1or Red Hat Enterprise Linux 5 1 with current maintenance updates for FICON FCP attachment to Linux on System z systems can be done through a switched fabric or direct attachment configuration For more information see Chapter 6 Linux considerations on page 81 12 2 Operating system prerequisites and enhancements The following minimum software levels are required to support the DS8800 z OS V1 8 z VM V5 4 IBM z VSE V4 2 Transaction Processing Facility TPF V4 1 with PTF Novell SUSE Linux Enterprise Server10 SP1 Red Hat Enterprise Linux 5 1 vvvvvy Certain functions of the DS8000 require later software levels than the minimum levels listed here See the IBM System Storage Interoperation Center SSIC at the following website http www ibm com systems support storage ssic interoperability wss Tip In addition to the IBM System Storage Interoperation Center SSIC see the Preventive Service Planning PSP bucket of the 2107 devices for software updates The PSP information can be found in the IBM Resource Link website http www ibm com servers resourcel ink svc03100 nsf 0penDatabase You need to register for an IBM registration ID IBM ID before you can sign in to the website Under the Planning section go to Tools to download the relevant PSP package
335. ou need to include it manually Tip The initRAMFS was introduced many years ago and replaced the initial RAM disk initrd The initrd reference is still being used however initRAMFS replaced initrd SUSE Linux Enterprise Server Kernel modules that must be included in the initRAMFS are listed in the etc sysconfig kerne file in the line that starts with INITRD_MODULES The order in which they are shown in this line is the order in which they are loaded at system startup Example 6 6 Example 6 6 Configuring SLES to include a kernel module in the initRAMFS x36501ab9 cat etc sysconfig kernel This variable contains the list of modules to be added to the initial Chapter 6 Linux considerations 91 ramdisk by calling the script mkinitrd like drivers for scsi controllers for Ivm or reiserfs INITRD MODULES thermal aacraid ata_piix processor fan jbd ext3 edd qla2xxx After adding the HBA driver module name to the configuration file rebuild the initRAMFS with the mkinitrd command This command creates and installs the image file with standard settings and to standard locations as shown in Example 6 7 Example 6 7 Creating the initRAMFS x36501ab9 mkinitrd Kernel image boot vmlinuz 2 6 32 12 0 7 default Initrd image boot initrd 2 6 32 12 0 7 default Root device dev disk by id scsi SServeRA_Drive_1_2D0DE908 partl dev sdal Resume device dev disk by id scsi SServeRA Drive _1 2DO0DE908 part3 de
336. p vg itso fc now active x36501ab9 mount dev vg_itso_fc lv_itso mnt lv_fc_itso x36501ab9 mount dev mapper vg_ds8000 1v_itso on mnt lv_itso type ext3 rw dev mapper vg_itso_fc lv_itso on mnt lv_fc_itso type ext3 rw 6 5 Troubleshooting and monitoring 114 This section provides information about the following areas of troubleshooting and monitoring Checking SCSI devices alternate methods Monitoring performance with the iostat command Working with generic SCSI tools Booting Linux from DS8000 volumes Configuring the QLogic BIOS to boot from a DS8000 volume Understanding the OS loader considerations for other platforms Installing SLES11 SP1 on a DS8000 volume vvvvvvy IBM System Storage DS8000 Host Attachment and Interoperability 6 5 1 Checking SCSI devices alternate methods The Linux kernel maintains a list of all attached SCSI devices in the proc pseudo file system The file proc scsi scsi contains basically the same information apart from the device node as the 1sscsi output It is always available even if 1sscsi is not installed To view the list as shown in Example 6 58 Example 6 58 An alternative list of attached SCSI devices x36501ab9 cat proc scsi scsi Attached devices Host scsi0 Channel 00 Id 00 Lun 00 Vendor ServeRA Model Drive 1 Rev V1 0 Type Direct Access ANSI SCSI revision 02 Host scsi2 Channel 00 Id 00 Lun 00 Vendor MATSHITA Model UJDA770 DVD CDRW Rev 1 24 Type
337. p is protected This is disk pool 1 of 1 disk pools that you selected to work with To add parity protected disk units to disk pool Ds_iasp click Add Parity Protected Disks To add pairs of disk units to be mirrored click Add Disks to be Mirrored To remove select the disk unit or units and click Remove NOTE Ifthe disk unit you are selecting to remove is setto be mirrored you will need to select them in pairs of equal capacity Selected disk units Add Disks to be Mirrored Add ileal let Disks Figure 11 19 Add disks to disk pool Chapter 11 IBM i considerations 231 10 Highlight the disks you want to add to the disk pool and click Add when a list of non configured units appears as shown in Figure 11 20 To add parity protected disk units to disk pool Ds_iasp select the disk unit or units and click Add Click Help for more information on why some disk units are not eligible Available disk units Figure 11 20 Select the disks to add to the disk pool 11 In the Disk Pool New Disk Pool Add Disk Units window Figure 11 21 click Next Disk pool Ds_iasp is protected This is disk pool 1 of 1 disk pools that you selected to work with To add parity protected disk units to disk pool Ds_iasp click Add Parity Protected Disks To add pairs of disk units to be mirrored click Add Disks to be Mirrored To remove select the disk unit or units and click Remove NOTE Ifthe disk unit you are selecting to remove
338. pacities the UCB s 64 K limitation is approaching at a rapid rate Therefore large volume support must be planned for the needs of installations that require super large volumes IBM has developed a volume named Extended Address Volume EAV recognized by host as 3390 Model A 3390A Full support for EAV volume was provided starting from z Os V1R12 Previous levels can have the same restrictions on sequential basic large partitioned PDS PDSE catalogs and BDAM data sets in the extended addressing space EAS 254 IBM System Storage DS8000 Host Attachment and Interoperability Support is enhanced to expand volumes to 65 520 cylinders using existing 16 bit cylinder addressing It is often referred to as 64 K cylinder volumes Components and products that previously shipped with 32 760 cylinders now also support 65 520 cylinders for example gt Direct access device space management DADSM or common volume table of contents VTOC access facility CVAF gt DFSMS functional component DFSMSdss gt ICKDSF a device support facility gt Data Facility Sort DFSORT a high speed data processing utility Checkpoint restart processing now supports a checkpoint data set that resides partially or wholly above the 32 760 cylinder boundary With logical volumes LVs the VTOC has the potential to grow large Callers such as DFSMSdss need to read the entire VTOC to find the last allocated data set control block DSCB In cases wher
339. pe Sun parameter When making the volume groups use the parameter type scsimap256 For native HBA drivers using the sd stack with the kernel drv sd conf file it is best to use persistent binding If you do not use persistent binding Solaris can assign a different SCSI device identifier SCSI ID other than the SCSI ID it used previously each time it rescans the bus for example in search for new devices Without persistent binding and changing SCSI IDs you would need to reconfigure applications or even the operating system which is in most cases an undesired behavior The methods of enabling persistent binding differ depending on your HBA The BM System Storage DS8000 Host Systems Attachment Guide SC26 7917 contains HBA settings for each supported type After creating the hostconnection objects in the DS8000 and connecting them to a volume group containing LUNs the storage system is providing them across the SAN allowing the Solaris server to scan the SCSI bus for new devices 9 4 Multipathing in Solaris As with other operating systems you typically use multiple paths to access a LUN from your Solaris server on the DS8000 system Multiple paths can help maximize the reliability and performance of your operating environment The DS8000 supports three multipathing technologies on Solaris gt IBM provides the System Storage Multipath SDD as part of the DS8000 at no extra charge gt Oracle Solaris has a native multipathing softw
340. pic iphdx 550_m50_landing htm IBM System Storage DS8000 Host Attachment and Interoperability gt IBM Power Blade servers http www ibm com systems power hardware blades index htm gt IBM i and System i Information Center http publib boulder ibm com iseries gt IBM Support Portal http www ibm com support entry portal gt System Storage Interoperation Center SSIC http www 03 ibm com systems support storage ssic interoperability wss gt BladeCenter Interoperability Guide BIG accessible on the following website http www 947 ibm com support entry portal docdisplay Indocid MIGR 5073016 gt To check for the latest IBM i program temporary fix PTF http www ibm com support entry portal 11 2 Using Fibre Channel adapters This section presents an overview of currently supported adapters in POWER systems or POWER based BladeCenters when connecting IBM i to a DS8000 Also provided in this section is the number of LUNs that you can assign to an adapter and the queue depth on a LUN with SCSI command tag queuing For more detailed up to date information see the websites listed in 11 1 4 Useful websites on page 204 Tip A Fibre Channel adapter in IBM i is also known as an input output adapter IOA 11 2 1 Native attachment The following Fibre Channel adapters are supported in IBM i system to connect DS8000 series in native mode 2787 PCI X Fibre Channel disk controller 5760 4 GB Fibre Chan
341. ping SDD devices to Windows drive letters 0 000 e eens 21 3 3 Clustering for Windows 2003 Server 00 cece 21 3 3 RETErENCOS siete Gad ate oS ERE ere GAs oe Rte ee AT Se Malian tee Rare eas 22 3 3 2 SDD SUDPOM sc dst kere dec ead Pee ares ane ideas a oe beaches A heen eee G 22 3 4 Using Multipath Input Output for Windows 2003 and 2008 20000 22 3 5 Partition alignment esr aaea a AEE RA E eee 23 3 6 Installing and configuring SDDDSM in Windows 2003 2008 05 23 3 6 1 SDDDSM for DS8000 2 2 eee 23 3 6 2 SDDDSM datapath query nauan 25 3 6 3 Windows 2008 and SDDDSM nananana aaea 26 3 7 Expanding dynamic disk for Windows 2003 Windows 2008 27 3 8 SAN boot SUPPO ioa na e a teens 32 3 9 Windows Server 2003 Virtual Disk Service support s s sasaaa anaana 32 3 9 1 VDS integration with DS8000 storage subsystems 200000505 33 3 9 2 Volume Shadow Copy Service 0000 eee 33 3 9 3 Required components 0 00 ete 34 Copyright IBM Corp 2011 All rights reserved iii iv 3 10 Hyper V considerations 00 0000 ce eee 36 3 10 1 Hyper V introduction 0 0 cee 36 3 10 2 Storage concepts for virtual machines 000 eee 37 3 10 3 Assigning a VHD to a virtual machine 0 000 ee 39 3 10 4 Assigning a pass through disk to a virtual machine 42 3 10 5 Cluster Share
342. policy value description user_settable PCM friend sddpcm PCM True none Reserve Key True load_balance Algorithm True no Device CLEARS its Queue on error True 0 Distributed Error Percentage True 50 Distributed Error Sample Time True 60 Health Check Interval True nonactive Health Check Mode True Location Label True 0x4090400000000000 Logical Unit Number ID False yes Support SCSI LUN reset True 0x100000 Maximum TRANSFER Size True 0x500507630af fc29f FC Node Name False 00f60dbc813b7cbd0000000000000000 Physical volume identifier False yes Use QERR bit True simple Queuing TYPE True 2 delay in seconds for SCSI TASK SET FULL True 20 Queue DEPTH True no_reserve Reserve Policy True 56 IBM System Storage DS8000 Host Attachment and Interoperability retry_timeout rw_timeout scbsy_dly scsi_id start_timeout unique_id ww_name 120 Retry Timeout True 60 READ WRITE time out value True 20 delay in seconds for SCSI BUSY True 0x201400 SCSI ID False 180 START unit time out value True 200B75TV181900007210790003IBMfcp Device Unique Identification False 0x500507630a4b429Ff FC World Wide Name False When using a redundant dual VIOS setup with DS8000 devices as virtual SCSI backing devices and native AIX MPIO in the client partition see the setup advice for virtual SCSI redundancy in PowerVM Virtualization on IBM System p Introduction and Configuration Fourth Edition SG24 7940 Figure 5 2 on page 67 shows an overview of preferred MPIO settings in
343. ps 9155 50 104 hmc wcl T2b3d Virtual Fibre Channel Adapter Properties ITSO LPAR 5 Virtual Fibre Channel adapter Adapter us Type of adapter Client WWPNs c050760293780034 c050760293780035 This adapter is required for partition activation Server partition v ITSO VIO 1 5 Server adapter ID 15 OK Cancel Help x Figure 5 1 WWPNs of a virtual Fibre Channel adapter Multipath drivers such as SDD or SDDPCM provide commands such as the pcmpath command for SDDPCM and the datapath command for SDD These commands can be used to manage and list DS8000 devices and their paths on AIX and to associate them with their DS8000 serial numbers 64 IBM System Storage DS8000 Host Attachment and Interoperability 5 2 Attaching virtual SCSI When using virtual SCSI emulation on the VIOS with DS8000 storage system volumes all the DS8000 volumes that are used as backing devices for the client partitions need to be mapped to the VIOS partition using the WWPNs of the physical Fibre Channel adapters in the VIOS partition as described in Using VSCSI with IBM VIOS and DS8000 on page 49 The virtual SCSI disks that are backed by DS8000 volumes on the VIOS appear as generic virtual SCSI disk devices on the AIX client partition without any storage system specific attributes as shown in Example 5 5 Example 5 5 DS8000 backed virtual SCSI disks on an AIX client partition Isdev Cc disk hdiskO Availa
344. r S E IDE Controller 0 Hard Drive Windows 2003 vhd 3 IDE Controller 1 DVD Drive est iso SCSI Controller 4 Network Adapter Figure 3 16 Add hard drive To remove the virtual storage adapter from this virtual machine click Remove Controller All virtual hard disks attached to this controller will be removed but not deleted Chapter 3 Windows considerations 39 3 To add a new hard drive click New as shown in Figure 3 17 Ca Hard Drive You can change how this virtual hard disk is attached to the virtual machine IF an operating system is installed on this disk changing the attachment might prevent the virtual machine From starting Controller Location scst Controller il fo in use S Media You can compact or convert a virtual hard disk by editing the vhd file Specify the Full path to the file Virtual hard disk vhd file Edit Inspect Browse Physical hard disk Disk 3 20 00 GB Bus 0 Lun 2 Target 0 TF the physical hard disk you want to use is not listed make sure that the disk is offline Use Disk Management on the physical computer to manage physical hard disks Figure 3 17 New hard drive properties 4 The New Virtual Hard Disk wizard is launched Click Next on the initial panel Select the type of virtual hard disk and click Next the available hard disk types are shown in Figure 3 18 a Fixed size The spac
345. r 3 Windows considerations 23 Figure 3 4 shows three disks connected by two paths to the server Computer Management ioj x m File Action view Window Help al xl e Amea Rema m Computer Management Local K System Tools E Event viewer Shared Folders Local Users and Groups g Performance Logs and Alert H E Device Manager Storage Ep Removable Storage Disk Defragmenter 2 3 Disk Management ie Services and Applications ts F LE i x346 TIC 3 Computer a WE es lt IBM 2107900 Multi Path Disk Device s IBM 2107900 Multi Path Disk Device lt IBM 2107900 Multi Path Disk Device lt IBM 2107900 SCSI Disk Device lt IBM 2107900 SCSI Disk Device lt IBM 2107900 SCSI Disk Device lt IBM 2107900 SCSI Disk Device lt IBM 2107900 SCSI Disk Device lt IBM 2107900 SCSI Disk Device lt IBM ServeRAID SCSI Disk Device lt IBM ServeRAID SCSI Disk Device E Display adapters Figure 3 4 Connecting SDDDSM devices in Windows Device Manager view In Figure 3 4 the three IBM 2107900 Multi Path Disk Devices are shown as real disks in Windows Device Manager under Disk drives The IBM 2107900 SCSI Disk Device is hidden by SDDDSM Figure 3 5 shows initialized disks in the Disk Management view LI Compute anage jal x Fie Action view Wi
346. r D LOOP ID 1 8 is Installed As Drive C Device Device Adapter Port Lun Vendor Product Product Number Type Number ID Number ID ID Revision 86 Disk a acadea A IBM 2107988 312 ROM BIOS Installed Figure 2 7 QLogic BIOS enabled messages Booting Emulex HBAs also support booting from SAN disk devices Press Alt E or Ctrl E to enable and configure the Emulex BIOS extension when the HBAs are initialized during server startup For more detailed instructions see the following Emulex publications gt Supercharge Booting Servers Directly from a Storage Area Network http www emulex com artifacts fc0b92e5 4e75 4f03 9 f0b 763811f47823 booting ServersDirectly pdf gt Enabling Emulex Boot from SAN on IBM BladeCenter http www emulex com artifacts 4f6391dc 32bd 43ae bcf0 1f51cc863145 enabl in g_boot_ibm pdf 2 2 2 Next steps For selected operating systems see the BM System Storage DS8000 Host Systems Attachment Guide SC26 7917 This guide provides procedures that are needed to boot a host off the DS8000 For information about identifying an optimal configuration and how to boot from multipathing devices see the IBM System Storage Multipath Subsystem Device Driver User s Guide GC52 1309 You can download this guide from the following FTP site ftp ftp software ibm com storage subsystem UG 1 8 3 0 2 3 Additional supported configurations For instances where the configuration that you want is not covered by th
347. rage pools and volume IDs from LSS of 60 per rank group 0 and LSS 61 per rank group 1 4 Assign DS8000 volume group to DS8000 host definition for the AIX server using the managehostconnect command if not already done with mkhostconnect command 62 IBM System Storage DS8000 Host Attachment and Interoperability Example 5 3 illustrates steps 2 through 4 on page 62 Example 5 3 Creating a host attachment and assigning DS8000 volumes to an AIX server dscli gt mkvolgrp hosttype pseries LPAR_1 Residency CMUC00030I mkvolgrp Volume group V16 successfully created dscli gt mkhostconnect wwname 10000000C9808618 hosttype pseries portgrp 45 p770_02_lpar1_fc0 CMUCO0012I mkhostconnect Host connection 001B successfully created dscli gt mkhostconnect wwname 10000000C9808619 hosttype pseries portgrp 45 p770_02_lpar1_fcl CMUCO0012I mkhostconnect Host connection 001C successfully created dscli gt mkfbvol cap 50 extpool p8 name p770_vol volgrp v45 6200 6201 CMUCO0025I mkfbvol FB volume 6200 successfully created CMUCO0025I mkfbvol FB volume 6201 successfully created dscli gt managehostconnect volgrp v45 45 CMUCOO016I managehostconnect Port group number 45 successfully modified To quickly check which volumes are assigned to a given AIX server on a DS8000 storage system use the DS8000 Command Line Interface DSCLI commands 1shostconnect and 1sfbvol as shown in Example 5 4 The Ishostconnect login command on DSCLI also sho
348. re Actions Refresh Rescan Disks Create VHD Attach YHD Figure 3 21 Create VHD Chapter 3 Windows considerations 41 2 Specify location and filename type and size 3 The new VHD will then be listed in disk management of the Windows Server 2008 host See Figure 3 22 Disk 0 Basic System Reserved c 556 93 GB 100 MB NTFS 556 83 Online Healthy System Active Primary Partition Healthy CaDisk 1 Unknown 2 00 GB 2 00 GB Not Initialized Unallocated Figure 3 22 VHD in disk management This feature provides an option for the Windows Server 2008 to benefit from advanced functions of VHDs such as snapshots or dynamic expansion 3 10 4 Assigning a pass through disk to a virtual machine 42 The following steps are required to attach a physical disk or LUN as a pass through disk to a virtual machine 1 Inthe disk management of the Windows 2008 server ensure that the LUN is visible and in state offline as shown in Figure 3 23 Re Server Manager WIN GD3O7VIJHE1 EA AAE a aime AE E EE gb ee vohme Layout Type Fie System status i El m Diagnostics Ga C Simple Basic NTFS Healthy Boot Page File Crash Dump Primary ati Configuration C passthrough F Simple Basic NTFS Healthy Primary Partition E SS Storage CaSystem Reserved Simple Basic NTFS Healthy System Active Primary Partition yb windows Server Backup i
349. releases Limited number of devices that can be attached Gaps in LUN sequence leading to incomplete device discovery Limited dynamic attachment of devices Non persistent device naming that can lead to re ordering No native multipathing YYYY Y Dynamic generation of device nodes Linux uses special files also called device nodes or special device files for access to devices In earlier versions these files were created statically during installation The creators of a Linux distribution tried to anticipate all devices that can be used for a system and created the nodes for them This often led to a confusing number of existing nodes or missing nodes In recent versions of Linux two new subsystems were introduced hotplug and udev Hotplug has the task to detect and register newly attached devices without user intervention and udev dynamically creates the required device nodes for them according to predefined rules In addition the range of major and minor numbers the representatives of devices in the kernel space was increased and they are now dynamically assigned With these improvements the required device nodes exist immediately after a device is detected and only the device nodes needed are defined Persistent device naming As mentioned earlier udev follows predefined rules when it creates the device nodes for new devices These rules are used to define device node names that relate to certain device characteristics In the c
350. resources These Web sites are also relevant as further information sources gt Documentation for the DS8000 http publib boulder ibm com infocenter dsichelp ds8000ic index jsp gt IBM data storage feature activation DSFA website http www ibm com storage dsfa gt System Storage Interoperation Center SSIC http www 03 ibm com systems support storage ssic interoperability wss Help from IBM IBM Support and downloads ibm com support IBM Global Services ibm com services 286 IBM System Storage DS8000 Host Attachment and Interoperability IBM System Storage DS8000 Host Attachment and Interoperability 0 5 spine 0 475 lt gt 0 873 Redbooks 250 lt gt 459 pages IBM System Storage DS8000 Host Attachment and Interoperability Learn how to attach DS8000 to open systems IBM System z and IBM i See how to gain maximum availability with multipathing Discover best practices and considerations for SAN boot This IBM Redbooks publication addresses host attachment and interoperability considerations for the IBM System Storage DS8000 series Within this book you can find information about the most popular host operating systems platforms including Windows IBM AIX VIOS Linux Solaris HP UX VMware Apple and IBM z 0S The topics covered in this book target administrators or other technical personnel with a working knowledge of storage systems and a general understandin
351. resources for use with IBM i use the Disk Magic tool to model the response times and utilization values Also apply the sizing guidelines such as the number of DS8000 ranks the number of IBM i adapters size of LUNs and so on It is best to apply the sizing guidelines before modelling with Disk Magic so that you can specify the values obtained by sizing such as number of ranks size of LUNs and so on when you start modelling your DS8000 In order to model with Disk Magic and to apply the sizing before modelling you need to collect performance data on IBM i The data must be collected with IBM i Collection Services in 5 minutes intervals for a few consecutive days After the data is collected use Performance tools licensed product number 5761 PT1 to create the following reports to use for sizing and Disk Magic modelling gt System report sections Disk utilization gt Resource report section Disk utilization gt Component report section Disk activity These reports give size for the peaks in I O per second IOPS writes per second and MBps Chapter 11 IBM i considerations 207 11 3 1 Sizing for natively connected and VIOS connected DS8000 The IBM i capability skip operations is the capability of efficiently reading or writing a non contiguous block of data An example of skip operations is a block of data that spans over 128 KB pages and some of the 4 KB pages in the 128 KB pages are not populated with data The skip opera
352. ric Logical Unit Access ALUA capable storage array gt PSP The default PSP for VMW_SATP_ALUA is MRU most recently used Tip The interoperability matrix at vmware com will list Fixed as default PSP for DS8000 However the code will automatically choose MRU Both PSP values are supported for DS8000 144 IBM System Storage DS8000 Host Attachment and Interoperability In order to better utilize all available paths to the DS8000 users should switch to RR round robin To do it select the ESXi host and perform the following steps 1 Go to Configuration Storage Adapters select one of the Fibre Channel HBAs connected to the DS8000 right click one of the LUNs listed and select Manage Paths as outlined in Figure 7 43 9 155 113 69 VMware ESXi 5 0 0 623860 Evaluation 45 days remaining STs ae Ae eel AT e U Configuration ESSE aa Me ue eee AGTER E Hardware Storage Adapters Processors Device Type WWN M vmhba32 Block SCSI lemory erage ISP81xx based 10 GbE FCoE to PCIExpress CNA Networkdrs E vmhba6 Fibre Channel 20 00 00 c0 dd 18 3b 79 21 00 00 c0 dd 18 3b 79 Since Minna vmhba7 Fibre Channel 20 00 00 c0 dd 18 3b 7b 21 00 00 c0 dd 18 3b 7b t io b 8Gb Fibre Channelto PCI Express HBA ae tome Fibre Channel 20 00 00 24 ff 2d c3 40 21 00 00 24 ff 2d c3 40 Advanced Settings Fibre Channel 20 00 00 24 ff 2d c3 41 21 00 00 24 ff 2d c3 41 SS Sa vmhba4 Fibre Channel 20 00 00
353. ring an initial program load or vary on a message is sent to the System Operator message queue QSYSOPR If a connection is missing and you confirm that the connection was removed you can update hardware service manager HSM to remove that resource The HSM tool can be used to display and work with system hardware from both a logical and a packaging viewpoint This tool also aids in debugging I O processors and devices and fixing failing or missing hardware You can access HSM in SST and DST by selecting the option to start a service tool 11 6 Configuration guidelines This section contains information about configuring DS8000 for IBM i connecting through a SAN connecting through a VIOS and for defining the DS8000 volumes to IBM i IBM i specific This section does not provide step by step procedures on how to set up the DS8000 for IBM i just the tasks that are specific or different when connecting IBM i 11 6 1 Creating extent pools for IBM i LUNs Create two large extent pools for IBM i LUNs each of them assigned to one DS8000 server with the rank group of 0 for workload balancing When creating the LUNs consider defining them in rotate extents mode also known as storage pool striping 11 6 2 Defining the LUNs for IBM i 220 For DS8000 natively attached to IBM i or attached through VIOS NPIV select the following GUI options when defining LUNs for IBM i gt Volume type iSeries protected or iSeries unprotected For mor
354. ring detection 8 3 Installing the ATTO software The following sections examine the installation of the ATTO driver software for OS X and the ATTO configuration tool The procedure for updating the flash firmware on the ATTO card is also covered 8 3 1 ATTO OS X driver installation 152 The DS8000 supported driver does not ship with ATTO products it must be downloaded 1 Download the supported Mac OS X Multipath Director Driver file from ATTO directly Look for the driver for IBM DS Series from the Fibre Channel HBAs for Storage Partners page On the HBA Vendor webpage click Solutions gt IBM solutions gt DS Solutions to get to the DS specific main page Get the needed installer s files by selecting your HBA type and login to the download page Again you need to reference the SSIC page at IBM to find supported driver levels IBM System Storage DS8000 Host Attachment and Interoperability 2 Important The Multipath Director Driver software is not included with the Host Bus Adapter product you need to download it from ATTO Install the HBA software For more information see the readme file that should be included When the software driver is installed successfully and the final OK button is clicked the server will need to be restarted 3 Verify that the ATTO adapter driver module is loaded into memory using the host system kextstat command as shown in Example 8 1 Example 8 1 Using kextstat to verify driv
355. rovide the information for the LUN that the machine loader uses to start the operating system separately The z VM system provides the set loaddev and query loaddev cp commands for this purpose as shown in Example 6 60 Example 6 60 Set and query SCSI IPL device in z VM SET LOADDEV PORTNAME 50050763 0a00029f LUN 40494000 00000000 QUERY LOADDEV PORTNAME 50050763 OA00029F LUN 40494000 00000000 BOOTPROG 0 BR_LBA 00000000 00000000 The port name provided is the DS8000 host port that is used to access the boot volume After the load device is set use the IPL program with the device number of the Fibre Channel HBA that connects to the DS8000 port and LUN to boot from Automate the IPL by adding the required commands to the z VM profile of the virtual machine 118 IBM System Storage DS8000 Host Attachment and Interoperability 6 5 7 Installing SLES11 SP1 on a DS8000 volume With recent Linux distributions the installation on a DS8000 volume is as easy as the installation on a local disk For installation keep in mind these additional considerations gt Identifying the right DS8000 volumes to install on gt Enabling multipathing during installation Tip After the SLES11 installation program YAST is running the installation is mostly hardware platform independent It works the same regardless of running on an x86 IBM Power System or System z server Start the installation process by booting from an installation DVD and follow the ins
356. rrors with HP UX If a Fibre Channel link from the HP UX system to the DS8000 fails native multipathing will automatically take the path offline It should also automatically bring the path back online after itis established again Example 10 9 shows the messages that are posted to the syslog when a link goes offline and then comes back online On HP UX 1 1iv3 this logfile is commonly found as var adm syslog syslog log and it contains entries from all subsystems so parsing through it might take a while The keywords to look for are Link Dead offline and online Example 10 9 Entries of syslog log by HP s native multipathing solution for Fibre Channel link failures May 30 12 38 03 rx6600 1 vmunix 0 3 1 0 Fibre Channel Driver received Link Dead Notification May 30 12 38 03 rx6600 1 vmunix class tgtpath instance 3 May 30 12 38 03 rx6600 1 vmunix Target path class tgtpath instance 3 has gone offline The target path h w path is 0 3 1 0 0x5005076303010143 MEE Later that same minute May 30 12 38 08 rx6600 1 vmunix Target path class tgtpath instance 4 has gone online The target path h w path is 0 3 1 0 0x5005076303080143 May 30 12 38 08 rx6600 1 vmunix class tgtpath instance 5 A system that is experiencing I O paths frequently dropping offline and coming back online will generate messages in the syslog file similar to those in Example 10 10 Example 10 10 Syslog entries for a large number of I O errors May 30 12 3
357. rs in the VIOS partition The VIOS does not have access to the client volumes The LUN is displayed in the client partition with the same storage system specific properties as with physical Fibre Channel attachment and the supported storage system specific multipathing driver must be installed on the client LPAR NPIV For IBM i Power Systems NPIV requires IBM i LIC 6 1 1 or later with exceptions as detailed in the following links NPIV capable Fibre Channel adapters and SAN switches are also required For a quick reference click the following link http www 03 ibm com support techdocs atsmastr nsf WebIndex PRS4563 For further details see the following website http www 03 ibm com systems support storage ssic interoperability wss The virtual Ethernet function also shown in Figure 4 1 on page 48 is basically provided by the IBM POWER Hypervisor enabling secure communication between logical partitions without the need for a physical I O adapter or cabling With shared Ethernet adapters on the VIOS the internal network traffic from the virtual Ethernet networks can be bridged out to physical Ethernet networks using a physical network adapter in the VIOS partition The storage virtualization capabilities by PowerVM and the VIOS are supported by the DS8000 series using DS8000 LUNs as VSCSI backing devices in the VIOS It is also possible to attach DS8000 LUNs directly to the client LPARS using virtual Fibre Channel adapters though the
358. rver cash write Upon connecting to the DS8k you will see this appear under the Device Manager view These two disks are connected by four paths to the server The IBM 2107900 Small Computer System Interface SCSI disk device is hidden by the SDD See Figure 3 2 1150 51008 WY Computer Disk drives E Display adapters i 2 DVD CD ROM drives s IBM 2107900 SCSI Disk Device lt IBM 2107900 SCSI Disk Device lt IBM 2107900 SCSI Disk Device IBM 2107900 SCSI Disk Device lt IBM 2107900 SCSI Disk Device s IBM 2107900 SCSI Disk Device lt IBM 2107900 SCSI Disk Device lt IBM 2107900 SCSI Disk Device IBM 2107900 SDD Disk Device IBM 2107900 SDD Disk Device s IBM PSG DDYS T36950M M SCSI Disk Device lt IBM PSG DDYS T36950M M SCSI Disk Device Figure 3 2 Device Manager view with SDD paths Chapter 3 Windows considerations 19 Figure 3 3 shows two SDD devices with the Disk Management view They can be allocated by initializing the disk drives and formatting the proper partition size i Computer Management m File Action View Window Help e Am2 A EH E Computer Management Local f volume Layout Type File System status Capacity Free Space Free Fault Toleral 5i H System Tools Sic Partition Basic NTFS Healthy System 33 89GB 26 45 GB 78 No fal Event viewer amp ds8000_1000 E Partition Basic NT
359. s This chapter provides the specifics of attaching the IBM System Storage DS8000 series system to System z hosts The following topics are covered gt Connectivity considerations Operating system prerequisites and enhancements Considerations for z OS DS8000 device definition and zDAC z OS FICON Discovery and Auto Configuration feature Extended Address Volume EAV support FICON specifics for a z OS environment z VM considerations VSE ESA and z VSE considerations I O Priority Manager for z OS TPC R V5 1 in a z OS environment Full Disk Encryption FDE Yv y vvvvvvy Copyright IBM Corp 2011 2013 All rights reserved 247 12 1 Connectivity considerations The DS8000 storage system connects to System z hosts using FICON channels with the addition of FCP connectivity for Linux for System z hosts For more information see the System z Connectivity Handbook SG24 5444 http www redbooks ibm com abstracts sg245444 html 12 1 1 FICON Check for dependencies on the host hardware driver level and the supported feature codes Your IBM service representative can help you determine your current hardware driver level on your mainframe processor complex For the IBM z9 and z10 servers FICON Express4 LX FC3321 at 10 km and FC3324 at 4 km and FICON Express4 SX FC3322 are available In addition z10 supports the new FICON Express8 LX FC3325 at 10 km and FICON Express8 SX FC3326 12 1 2 LINUX FCP connectivity Y
360. s complex and cheaper channel extender which only performs frame forwarding rather than a channel extender that dissects every channel control word CCW to optimize the transfer through the channel extender to get the best performance These are typically channel extenders that have extended remote copy XRC emulation running on them An enhancement was implemented on the standard FICON architecture Fibre Channel single byte command set 3 FC SB 3 layer with a new protocol for persistent IU pacing It was achieved for the DS8000 z OS Global Mirror XRC SDM read record set RRS data transfer from a DS8000 to the SDM host address space The control units that support the extended distance FICON feature can increase the pacing count The pacing count is the number of lUs that are inflight from a channel to a control unit IU pacing Standard FICON supports IUs pacing of 16 IUs in flight Extended Distance FICON now extends the IU pacing for the RRS CCW chain to permit 255 IUs inflight without waiting for an acknowledgement from the control unit eliminating engagement between the channel and control unit This support allows the channel to remember the last pacing information and use this information for subsequent operations to avoid performance degradation at the start of a new I O operation Extended Distance FICON reduces the need for channel extenders in DS8000 series 2 site and 3 site z OS Global Mirror configurations because of the increas
361. s os et ay ES 77 5 5 4 Impact of DS8000 storage pool striping 0 eee 77 5 6 Using AIX access methods for I O 1 0 eee 77 5 6 1 Synchronous VO se cise cele wed daeaie oe ee eed a eee Ses See RNAS 77 5 6 2 Asynchronous WOls ei ake ahaa eid eae had oe kaneis Feels ares 78 5 6 3 Concurrent I O ae con 2bae tine ade ceca oh ea oe guns vole deuce S 78 5 0 4 Direct Oe sewn Ges Aeneas atu ie tan anaes eb alas een HM ea hae eit onset 78 5 7 Expanding dynamic volume with AIX suasusu anaa aaaea 79 5 8 SAN DOOUSUPPOMT 38 bubs no ead are ee E Sleek lataa E daw aeae ated a Pala ees 80 Chapter 6 Linux considerations 00 000 c eee eee 81 6 1 Working with Linux and DS8000 0 82 6 1 1 How Linux differs from other operating systems 0000 eee eee 82 6 1 2 Attaching Linux server to DS8000 resources 0 000 eee eee 82 6 1 3 Understanding storage related improvements to Linux 85 6 2 Attaching to a basic NOSI oraora n irpan EE EREE ERE EE E EAR a ER RNA 86 6 2 1 Platform specific information saaa aaae 86 6 2 2 Configuring Fibre Channel attachment s a sasaaa aeaaaee 89 6 2 3 Determining the WWPN of installed HBAS 0 00000 e eee eae 93 6 2 4 Checking attached volumes 0 0000 e ete 93 6 2 5 Linux SCSI addressing 0 00 c cette 94 6 2 6 Identifying DS8000 devices 0 0 ccc tee 95 6 2 7 Adding DS8000 volumes to Linux on Sy
362. s providing the same DS8000 LUN as backing device to the same client LPAR the UDID of the DS8000 disk is used by AIX MPIO to properly identify all the paths for a given virtual SCSI disk The UDID of the Virtual SCSI disk can also be used on the client partition for identifying the volume ID of the DS8000 volume used as backing device as shown in 5 2 Attaching virtual SCSI on page 65 The UDID shown in Example 4 9 contains the DS8000 serial number volume ID and device type For example the unit device identifier UDID 200B75TV181900007210790003IBMfcp contains the DS8000 storage image serial number 75TV181 the DS8000 volume number 9000 and the device type 2107 900 Chapter 4 Virtual I O Server considerations 55 Because virtual SCSI connections operate at memory speed generally no added performance is gained by adding multiple adapters between a VIOS and the client partition However it is a best practice to use separate virtual SCSI adapter server and client pairs for boot or AIX rootvg disks and data disks in addition to backing devices from separate storage systems In general for AIX virtual I O client partitions each adapter pair can handle up to 85 virtual devices with the default queue depth of three For IBM i clients up to 16 virtual disks and 16 optical devices are supported Tip The queue depth on LUNs connected by virtual SCSI to IBM i is 32 In situations where virtual devices per partition are expected to e
363. s systems to open systems hosts It includes references to available documentation and additional information The following topics are covered gt Configuration resources Using the DS8000 as a boot device Additional supported configurations Multipathing support for Subsystem Device Drivers Yv y Copyright IBM Corp 2011 2013 All rights reserved 2 1 Configuration resources This section describes the online resources where you can find detailed and up to date information about supported configurations settings device driver versions and so on for open systems Because of the high innovation rate in the IT industry the support information is updated frequently Visit these resources regularly to check for updates 2 1 1 IBM resources Resources covered in this section include the IBM System Storage Interoperation Center SSIC host systems attachment installation and troubleshooting as provided by IBM IBM System Storage Interoperation Center For information about supported Fibre Channel FC host bus adapters HBAs and the required firmware and device driver levels for all IBM storage systems visit the IBM SSIC at the following website http www ibm com systems support storage ssic interoperability wss For each list box select a storage system a server model an operating system and an HBA type Each list box provides a list of all supported HBAs together with the required firmware and device driver leve
364. s to be set to Round Robin RR as described in 7 3 Multipathing on page 142 While there is no general advice on the number of paths to use various configuration maximums need to be observed to ensure scalability in the future See Table 7 3 Table 7 3 Configuration Maximums Item Maximum Number of paths to a LUN 32 Number of total paths per ESXi host 1024 Number of LUNs per ESXi host 256 If each LUN has four paths to the storage device the maximum number of 256 LUNs is possible With an increasing number of paths the maximum number of LUNs will decrease While a high number of paths per LUN increases reliability and throughput the possible number of LUNs is reduced RDMs will further decrease the number LUNs available for datastores Chapter 7 VMware vSphere considerations 149 150 IBM System Storage DS8000 Host Attachment and Interoperability Apple considerations This chapter provides information for the specifics of attaching IBM DS8000 storage system to host systems running Mac OS X The following topics are covered Configuring the Apple host on a DS8000 Installing the ATTO software Using the ATTO Configuration Tool Creating a file system on Apple Mac OS X Troubleshooting vvvvy Attention At the time of writing this book no 8Gb s Host Bus Adapters HBAs were officially supported by the DS8800 storage system The ATTO 8Gb HBA is awaiting certification by IBM and was used successfully
365. s to which DS8000 LUN evaluate the output from running the datapath query device command and from the Windows Disk Management window In Example 3 1 on page 20 if you listed the vpaths you can see that SDD DEV 0 has DEVICE NAME Disk2 You can also see the serial number of the disk is 75065711000 which is displayed as LUN Identifier 1000 on DS8000 serial 7506571 Then look at the Windows Disk Management window an example of which is shown in Figure 3 3 on page 20 In this example Disk 2 is Windows drive letter E circled in blue The Windows drive letter G circled in red corresponds to the SDD DEV Now that you have mapped the LUN ID to a Windows drive letter if drive letter E was no longer required on this Windows server you can safely unassign LUN ID 1000 on the DS8000 with serial number 7506571 with the understanding that you have removed the correct drive 3 3 Clustering for Windows 2003 Server SDD 1 6 0 0 or later is required to support load balancing in Windows clustering When running Windows clustering clustering failover might not occur when the last path is being removed from the shared resources Chapter 3 Windows considerations 21 3 3 1 References For additional information see the Microsoft article Removing the HBA cable on a server cluster at the following website http support microsoft com default aspx scid kb en us Q294173 You can check your SDD version and MSCS settings to reduce reservation
366. s updated A separate entry in the boot menu provides the ability to switch between the single and multipath mode when necessary See the Linux distribution specific documentation in 6 1 2 Attaching Linux server to DS8000 resources on page 82 for more detail Chapter 6 Linux considerations 121 122 The installer does not implement any device specific settings such as creating the etc multipath conf file You must do this manually after the installation See 6 2 9 Attaching DS8000 Considerations on page 104 for more information Because DM Multipath is already started during the processing of the InitRAMFS you must also build a new InitRAMFS image after changing the DM Multipath configuration See Initial Fibre Channel driver availability on page 91 Linux tips It is possible to add device mapper layers on top of DM Multipath such as software RAID or LVM The Linux installers support these options Red Hat Enterprise Linux 5 1 and later also supports multipathing for the installation You can enable this option by adding the mpath option to the kernel boot line of the installation system Anaconda the Red Hat installer then offers to install to multipath devices IBM System Storage DS8000 Host Attachment and Interoperability VMware vSphere considerations This chapter provides information about the technical specifications for attaching IBM System Storage DS8000 systems to host systems running VMware vSph
367. scli gt Isfbvol 4700 Name ID accstate datastate configstate deviceMTM datatype extpool cap 2 30B cap 10 9B cap blocks ITSO_p770_1_4700 4700 Online Normal Normal 2107 900 FB 512 P53 18 0 37748736 Isvg p dvevg dvevg PV_NAME PV STATE TOTAL PPs FREE PPs FREE DISTRIBUTION hdiskO active 286 5 00 00 00 00 05 Isvg 1 dvevg dvevg LV NAME TYPE LPs PPs PVs LV STATE MOUNT POINT dvelv jfs2 280 280 1 open syncd dvefs log1v00 jfs2log 1 1 1 open syncd N A Isfs dvefs Name Nodename Mount Pt VFS Size Options Auto Accounting dev dvelv dvefs jfs2 36700160 rw yes no If more space is required in this file system two options are available with the AIX operating system either add another DS8000 logical volume which is a physical volume on AIX to the AIX volume group or extend the DS8000 logical volume and subsequently adjust the AIX LVM definitions The second option is demonstrated in Example 5 22 The DSCLI is used to extend the DS8000 logical volume On the attached AIX host the configuration is changed with the cfgmgr and chvg AIX commands Afterwards the file system is expanded online and the results are displayed Example 5 22 Dynamic volume expansion of DS8000 logical volume and AIX file system dscli gt chfbvol cap 24 4700 CMUC00332W chfbvol Some host operating systems do not support changing the volume size Are you sure that you want to resize the volume y n y CMUC00026I chfbvol FB volume 4700 succe
368. sk by id and dev disk by path The nodes for DS8000 volumes in dev disk by id provide a unique identifier that is composed of the WWNN of the DS8000 system and the DS8000 volume ID as shown in Example 6 13 Example 6 13 The dev disk by id device nodes x36501ab9 1s 1 dev disk by id cut c 44 scsi 3600507630af fc29f0000000000004690 gt scsi 3600507630af fc29f0000000000004691 gt scsi 3600507630af fc29f0000000000004790 gt scsi 3600507630af fc29f0000000000004791 gt The udev subsystem already recognizes that more than one path exists to each DS8000 volume It creates only one node for each volume instead of four Important The device nodes in dev disk by id are persistent whereas the dev sdx nodes are not They can change when the hardware configuration changes Do not use dev sdx device nodes to mount file systems or specify system disks sdj sdk sdl sdq a SS e In dev disk by path there are nodes for all paths to all DS8000 volumes Here you can see the physical connection to the volumes starting with the Peripheral Component Interconnect PCI identifier of the HBAs through the remote port represented by the DS8000 WWPN to the LUN of the volumes as shown in Example 6 14 Example 6 14 The dev disk by path device nodes x36501ab9 pci 0000 pci 0000 pci 0000 pci 0000 pci 0000 pci 0000 pci 0000 pci 0000 pci 0000 pci 0000 pci 0000 pci 0000 pci
369. sk partition alignment terminology http support microsoft com kb 929491 http technet microsoft com en us 1ibrary dd758814 28v sq1 100 29 aspx 3 6 Installing and configuring SDDDSM in Windows 2003 2008 The Subsystem Device Driver Device Specific Module SDDDSM installation is a package for DS8000 devices on the Windows Server 2003 and Windows Server 2008 With MPIO SDDDSM supports the multipath configuration environments in the IBM System Storage DS8000 It resides in a host system with the native disk device driver and provides the following functions gt Enhanced data availability Dynamic I O load balancing across multiple paths Automatic path failover protection Concurrent download of licensed internal code Path selection policies for the host system vvvy Also be aware of the following limitations gt SDD is not supported on Windows 2008 gt Foran HBA driver SDDDSM requires the Storport version of the HBA Miniport driver SDDDSM can be downloaded from the following website http www ibm com support docview wss rs 540 amp context ST52G7 amp dc D430 amp ui d ssg1S 4000350 3 6 1 SDDDSM for DS8000 Ensure that SDDDSM is installed before adding additional paths to a device otherwise the operating system can lose the ability to access existing data on that device You will also see repeated disks For details see the IBM System Storage Multipath Subsystem Device Driver User s Guide SC30 4131 Chapte
370. solution called Hyper V Manager Hyper V uses the term partition for virtual machines The Hyper V Hyper Visor requires at least one parent partition which hosts Windows Server 2008 Inside this parent partition the virtualization software is running which controls access of the child partitions to the physical hardware The child partitions have virtual hardware components that are dynamically mapped through the virtual machine bus VM bus to the physical hardware by the parent partition IBM System Storage DS8000 Host Attachment and Interoperability Figure 3 13 shows the architecture of Hyper V Microsoft Hyper V server Virtualization Service Parent partition Child partition Child partition with Windows 2008 vol ERC Application Application i VM processes Kemel Kemel Virtualization Service Kemel Consumer Consumer Device drivers VSC VSC Virtualization Service Provider VSP Hardware Hardware Virtualization layer hypervisor Hardware Figure 3 13 Microsoft Hyper V architecture 3 10 2 Storage concepts for virtual machines Virtual machines consist of several files which can contain the data of one or more virtual hard drives saved memory saved device states and so on All these files are stored on NTFS partitions that are managed by the Windows Serv
371. splex scope Chapter 12 IBM System z considerations 251 gt Auto configuration For high availability reasons when zDAC proposes channel paths it looks at single points of failure only It does not consider any channel or port speed or any current performance information After a storage subsystem is explored the discovered information is compared against the target IODF paths are proposed to new control units and devices are displayed to the user With that scope of discovery and auto configuration the target work IODF is being updated When using ZDAC keep in mind the following considerations YYYY Y Physical planning is still your responsibility Logical configurations of the Storage Subsystem are still done by you Decide what z OS image should be allowed to use the new devices Determine how the new devices must be numbered Decide how many paths to the new control units need to be configured System z Discovery and Auto Configuration ZDAC Common Fabrics RNID for Topology Discovery NewFICON ELS for rapid discovery of CU images New FICON Channel Commands for issuing ELS commands to Name Server and Storage 1 Figure 12 1 zDAC concept For more information about zDAC see the Redbooks publication z OS V1R12 HCD User s Guide SC33 7988 Attention In z OS environments make sure that SAN switch zoning is enabled and configured properly before start to us
372. ssfully modified cfgmgr chvg g dvevg Isvg p dvevg dvevg PV_NAME PV STATE TOTAL PPs FREE PPs FREE DISTRIBUTION Chapter 5 AlX considerations 79 hdiskO active 382 101 00 00 00 24 77 chfs a size 45000000 dvefs Filesystem size changed to 45088768 Isvg p dvevg dvevg PV_NAME PV STATE TOTAL PPs FREE PPs FREE DISTRIBUTION hdisk0 active 382 37 00 00 00 00 37 Isvg 1 dvevg dvevg LV NAME TYPE LPs PPs PVs LV STATE MOUNT POINT dvelv jfs2 344 34 1 open syncd dvefs 1og1v00 jfs2log 1 1 1 open syncd N A Isfs dvefs Name Nodename Mount Pt VFS Size Options Auto Accounting dev dvelv dvefs jfs2 45088768 rw yes no For the online size extension of the AIX volume group you might need to deactivate and then reactivate the AIX volume group for LVM to see the size change on the disks See the AIX documentation for more information at this website http publib16 boulder ibm com pseries index htm Tip The g option in the chvg command was not valid for concurrent vgs or for rootvg That restriction was removed on AIX release 6 1 TL3 or later 5 8 SAN boot support The DS8000 is supported as a boot device on Power Systems servers that have Fibre Channel boot capability For additional information see the IBM System Storage DS8000 Host Systems Attachment Guide GC27 2298 02 80 IBM System Storage DS8000 Host Attachment and Interoperability Linux considerations This chapter presents the specifics
373. stands for Version 6 Release 1 Modification level 1 When specifying support for a certain function just by version and release it means that any modification level supports it However if support for a function needs a certain modification level that level is specified 202 IBM System Storage DS8000 Host Attachment and Interoperability From IBM i V7 1 on IBM i modification levels are replaced by a new release delivery mechanism called a Technology Refresh A Technology Refresh TR is delivered as a PTF Group for a base release When specifying support for a certain function just by version and release it means that any TR supports it However if support for a function needs a certain TR level that level is specified Here are some requirements gt IBM i software levels V5 4 V6 1 and V7 1 support native attachment of the DS8000 gt Connection of DS8000 to IBM i with VIOS NPIV requires IBM i V6 1 1 V7 1 gt Connection of DS8000 to IBM i with VIOS VSCSI requires IBM i level V6 1 or V7 1 VIOS levels For connecting DS8000 to IBM i with VIOS NPIV you need the following VIOS levels gt In POWER server VIOS V2 1 2 or higher gt In BladeCenter VIOS V2 2 or higher Attachment of IBM i with VIOS VSCSI requires VIOS V1 5 0 or higher Needed levels for multipath Multipath for native connection is supported on all IBM i levels that are listed to support native attaching When using VIOS VSCSI or NPIV the multipath is achieve
374. stem You can define up to eight connections from multiple I O adapters on an IBM i server to a single logical volume in the DS8000 Each connection for a multipath disk unit functions independently Several connections provide availability by allowing disk storage to be used even if a single path fails If you have a configuration where the logical units are only assigned to one I O adapter you can easily change to multipath by assigning the logical units in the DS8700 to another I O adapter The existing DDxxx drives change to DMPxxx and new DMPxxx resources are created for the new path It might be possible in your environment to re assign logical volumes to other I O adapters but careful planning and implementation are required Chapter 11 IBM i considerations 217 Multipath is important for IBM i because it provides greater resilience to SAN failures which can be critical for IBM i due to the single level storage architecture Multipath is not available for IBM i internal disk units but the possibility of path failure is much less with internal drives It is because there are fewer interference points where problems can occur such as long fiber cables and SAN switches increased possibility of human error when configuring switches and external storage and the concurrent maintenance on the DS8000 which can make certain paths temporarily unavailable Many IBM i clients still have their entire environment on the server auxiliary storage poo
375. stem Z 000 eee eee 96 6 2 8 Setting up device mapper multipathing 0 0 0 c eee eee ee 98 6 2 9 Attaching DS8000 Considerations 00 00 cee eee 104 IBM System Storage DS8000 Host Attachment and Interoperability 6 2 10 Understanding non disruptive actions on attached hosts 106 6 2 11 Adding new DS8000 host ports to Linux on System Z 0005 108 6 3 Resizing DS8000 volumes dynamically 0 0 0 cece eee 109 6 4 Using FlashCopy and remote replication targets 0 0 0 cee eee eee 111 6 4 1 Using a file system residing on a DS8000 volume 0000 00s 111 6 4 2 Using a file system residing in a logical volume managed by LVM 112 6 5 Troubleshooting and monitoring 00 c eet eee 114 6 5 1 Checking SCSI devices alternate methods 00 0 cece eee eee 115 6 5 2 Monitoring performance with the iostat command 000005 116 6 5 3 Working with generic SCSI tools 00 eee 116 6 5 4 Booting Linux from DS8000 volumes 0 00 e cee eee 116 6 5 5 Configuring the QLogic BIOS to boot from a DS8000 volume 117 6 5 6 Understanding the OS loader considerations for other platforms 118 6 5 7 Installing SLES11 SP1 on aDS8000 volume 000 00 e eee 119 6 5 8 Installing with YAST 0 0 0 00 cette ee 121 Chapter 7 VMware vSphere considerations 0000
376. stics for every volume in the storage subsystem These measurements include the count of the number of operations from DASD cache to the back end storage the number of random operations the number of sequential reads and sequential writes the time to execute those operations and the number of bytes transferred These statistics are placed in the SMF 74 subtype 5 record For more information see the Redbooks publication MVS System Management Facilities SMF SA22 7630 Migration considerations The DS8000 is supported as an IBM 2105 for z OS systems without the DFSMS and z OS small program enhancements SPEs installed It allows clients to roll the SPE to each system in a sysplex without having to take a sysplex wide outage You must perform an IPL to activate the DFSMS and z OS portions of this support Coexistence considerations IBM provides support for the DS8000 running in 2105 mode on systems that do not have this SPE installed The support consists of the recognition of the DS8000 real control unit type and device codes when it runs in 2105 emulation on these down level systems IODF files created by HCD can be shared on systems that do not have this SPE installed Additionally you can use existing IODF files that define IBM 2105 control unit records for a 2107 subsystem as long as 16 or fewer logical systems are configured in the DS8000 12 4 Extended Address Volume EAV support As storage facilities tend to expand to larger ca
377. successfully extended the volume DISKPART gt detail volume Disk Status Size Free Dyn Gpt Disk 3 Online 18 GB 0 B Readonly No Hidden No No Default Drive Letter No Shadow Copy No Figure 3 10 shows the Disk Manager view of the result for the volume expansion o Computer Management Eifel xi E File Action View window Help x e Am e m m Computer Management Local E System Tools E Event viewer Shared Folders Local Users and Groups z Performance Logs and Alert Device Manager Storage Ep Removable Storage Disk Defragmenter a Sic Partition Basic NTFS Healthy Boot 68 36 GB 62 00 GB Ste Partition Basic NTFS Healthy System 33 90GB 33 83 GB GIDS8000_EXP Z Partition Basic NTFS Healthy 17 99GB 17 93 GB aE FDisk Management Ta Services and Applications ZPDisk 2 Basic 12 00 GB 12 00 GB Online Unallocated 2SDisk 3 Basic DS8000_EXP 2 17 99 GB 17 99 GB NTFS Online Healthy 4 gt W Unallocated iE Primary partition Figure 3 10 Disk Manager after expansion in diskpart Chapter 3 Windows considerations 31 The disk storage expanded in this example is referred to as a basic disk Dynamic disks can be expanded by expanding the underlying DS8000 volume The new space is displayed as unallocated space at the end of the disk I
378. system is to use the Isscsi command as shown in Example 6 12 Example 6 12 Listing attached SCSI devices x36501ab9 Isscsi 0 0 0 0 disk ServeRA Drive 1 V1 0 dev sda 2 0 0 0 cd dvd MATSHITA UJDA770 DVD CDRW 1 24 dev sr0 3 0 0 0 disk IBM 2107900 288 dev sdb 3 0 0 1 disk IBM 2107900 288 dev sdc 3 0 0 2 disk IBM 2107900 288 dev sdd 3 0 0 3 disk IBM 2107900 288 dev sde 3 0 1 0 disk IBM 2107900 288 dev sdf 3 0 1 1 disk IBM 2107900 288 dev sdg 3 0 1 2 disk IBM 2107900 288 dev sdh 3 0 1 3 disk IBM 2107900 288 dev sdi 4 0 0 0 disk IBM 2107900 288 dev sdj 4 0 0 1 disk IBM 2107900 288 dev sdk 4 0 0 2 disk IBM 2107900 288 dev sdl 4 0 0 3 disk IBM 2107900 288 dev sdm 4 0 1 0 disk IBM 2107900 288 dev sdn 4 0 1 1 disk IBM 2107900 288 dev sdo 4 0 1 2 disk IBM 2107900 288 dev sdp 4 0 1 3 disk IBM 2107900 288 dev sdq Chapter 6 Linux considerations 93 Example 6 12 on page 93 shows that Linux recognized 16 DS8000 devices By viewing the SCSI addresses in the first column you can determine that there actually are four DS8000 volumes each connected through four paths Linux creates a SCSI disk device for each of the paths Red Hat Enterprise Linux installer The Red Hat Enterprise Linux installer does not install Isscsi by default It is shipped with the distribution but must be selected for installation 6 2 5 Linux SCSI addressing 94 The quadruple in the first column of t
379. t He is a DS8000 and XIV Top Gun Specialist and has good knowledge of Storage on IBM z OS environments with 25 years experience His areas of experience also include critical situation management for storage environments Ulrich Rendels is an IT Specialist working for IBM Germany He has been with IBM for 15 years starting as a Product Field Engineer for 7135 and ESS For the past 10 years he has been a member of world wide storage development test teams supporting ESSNI DS8000 Development and currently qualifying IBM SVC Storwize V7000 and Storwize Unified systems Mario Rodriguez is an IBM Certified Systems Expert working for IBM Uruguay since 2001 He holds a degree in Computer Science and also multiple vendor certifications from IBM Microsoft VMware LPI and Comptia His areas of expertise include SAN Switches Storage Systems AIX Linux and VMware His main role in IBM Uruguay is to provide technical support services for virtualization and storage products Many thanks to the people at the IBM Systems Lab Europe in Mainz Germany who helped with the equipment provisioning and preparation Uwe Heinrich Muller Gunter Schmitt Mike Schneider Dietmar Schniering Uwe Schweikhard IBM Systems Lab Europe Mainz Germany Special thanks to the Enterprise Disk team manager Bernd M ller and the ESCC director Klaus JUrgen Runger for their continuous interest and support regarding the ITSO Redbooks projects Thanks to the fol
380. t a specific kernel module x36501ab9 modinfo qla2xxx filename lib modules 2 6 32 12 0 7 default kernel drivers scsi qla2xxx qla2xxx ko version 8 03 01 06 11 1 k8 license GPL description QLogic Fibre Channel HBA Driver author QLogic Corporation depends scsi_mod scsi_transport_fc supported yes vermagic 2 6 32 12 0 7 default SMP mod_unload modversions parm ql2xlogintimeout Login timeout value in seconds int parm qlport_down_retry Maximum number of command retries to a port parm ql2xplogiabsentdevice Option to enable PLOGI to devices that 90 IBM System Storage DS8000 Host Attachment and Interoperability Tip The zfcp driver for Linux on System z automatically scans and registers the attached volumes only in the most recent Linux distributions and only if NPIV is used Otherwise you must configure the volumes for access The reason is that the Linux virtual machine might not use all volumes that are attached to the HBA See Linux on System z running in a virtual machine under z VM on page 88 and 6 2 7 Adding DS8000 volumes to Linux on System z on page 96 for more information Using the FC HBA driver during installation You can use DS8000 volumes attached to a Linux system during installation This provides the ability to install all or part of the system to the SAN attached volumes The Linux installers detect the FC HBAs load the necessary kernel modules scan for volumes and offer them in th
381. t configuration program The host supports 255 logical control units when the DS8000 is defined as UNIT 2107 You must install the appropriate software to support this setup If you do not have the required software support installed you can define the DS8000 as UNIT 2105 but then only the 16 logical control units LCUs of address group 0 can be used The Control Unit CU device is a 2107 242x for any DS8000 device type In HCD you can use 3990 as a control unit type for 3380 and 3390 devices However you must use 2107 242x as the control unit type for 3380B 3380A 3390B and 3390A devices For more detailed information about how to configure DS8000 in z OS environment see the Redbooks publication I O Configuration Using z OS HCD and HCM SG24 7804 and the z OS V1R12 HCD User s Guide SC33 7988 Defining additional subchannels Starting with z9 processors you can define an additional subchannel set with ID 1 class 1 SS 1 on top of the existing subchannel set class O SS 0 in a channel subsystem With this additional subchannel set you can configure more than 2 x 63 K devices for a channel subsystem With z OS V1 7 you can define PAV alias devices of type 3380A and 3390A of the DS8000 2107 DASD control units to SS 1 Device numbers can be duplicated across channel subsystems and subchannel sets For more information about subchannel configuration on the I O Definition File ODF and HCD see the Redpaper publication Multiple Subchan
382. t queue I Os so it depends on the release of SDD SDDPCM on the other hand does not queue IOs before sending them to the disk device driver The hdisks have a maximum number of in flight I Os that s specified by its queue_depth attribute And FC adapters also have a maximum number of in flight I Os specified by num_cmd_elems The disk subsystems themselves queue IOs and individual physical disks can accept multiple I O requests but only service one at a time For more information about how to tune the queue_depth at the different levels see the following website http www 03 ibm com support techdocs atsmastr nsf WebIndex 1D105745 5 3 2 timeout_policy attribute Starting from SDDPCM v2 6 3 0 the path selection and management algorithm has enhanced to reduce this I O performance degradation as described before A new device attribute timeout_policy is introduced This attribute provides several options to manage a path connected to an unstable FC port in different ways to avoid being toggled between Open and FAILED states frequently This feature is designed to reduce the I O performance degradation symptom caused by intermittently failing links which can result in paths cycling between the online and offline states Under this failing link condition I O fails with TIMEOUT error and the path was put to FAILED state then subsequently brought back online following the successful completion of a health check command The recovered p
383. t to the DS8000 gt PPRC A layout of which storage subsystems are involved in Copy Services relationships and the type of connectivity between them gt Encryption A layout of TKLM servers and their vital information as well as contact information for security administrators For future planning IBM can provide assistance in developing these types of documentation Contact your IBM support representative for more information about services offered 1 3 SAN considerations The following list contains some general considerations regarding maintenance and configuration for Storage Area Networks SANs These rules are valid for both open and z OS environments gt Periodically check the SAN from the SAN Administrator for errors Always do this operation before any hardware change gt Fix any errors before making any configuration changes to avoid multiple problems in the SAN environment gt Before implementing any new configuration such as installing new hardware verify the firmware levels on all the equipment If needed update the firmware on all SAN switches to be sure that any old compatibility issue was resolved gt Before proceeding with any new hardware installation evaluate the SAN bandwidth to be sure that is not already to close to its limits This precaution can help avoid problems after the new hardware becomes operational If a client needs assistance in performing these checks IBM provides many levels of SA
384. tallation configuration panels until you reach the Installation Settings panel Figure 6 3 Using the GUI The Linux on System z installer does not automatically list the available disks for installation You will see a Configure Disks panel before you get to the Installation Settings where you can discover and attach the disks that are needed to install the system using a GUI At least one disk device is required to perform the installation To install SLES11 SP1 on a DS8000 volume follow these steps 1 Inthe Installation Settings panel Figure 6 3 click Partitioning to perform the configuration steps required to define the DS8000 volume as system disk Installation Settings Click any headline to make changes or use the Change menu below Overview Expert Keyboard Layout 5 Oot partition dev sde 100 00 GB with ext3 Use dev sda3 as swap Figure 6 3 SLESS11 SP1 installation settings Chapter 6 Linux considerations 119 2 In the Preparing Hard Disk Step 1 panel Figure 6 4 select the Custom Partitioning for experts button It does not matter which disk device is selected in the Available Disks field Then click Next lt s Preparing Hard Disk Step 1 Hard O sx Available Disks 1 dev sda 2 devisdb 3 devisde 4 jdavisat foe Custom Partitioning for experts Help Abort Back Next Figure 6 4 Preparing Hard Disk Step 1 3 In the E
385. tapath query device 1 Total Devices 2 DEV 0 DEVICE NAME Disk2 PartO TYPE 2107900 POLICY OPTIMIZED SERIAL 75065711000 LUN IDENTIFIER 6005076303FFC0B60000000000001000 20 IBM System Storage DS8000 Host Attachment and Interoperability Path Adapter Hard Disk State Mode Select Errors 0 Scsi Port2 BusO Disk2 Part0 OPEN NORMAL 22 0 1 Scsi Port2 Bus0 Disk2 Part0 OPEN NORMAL 32 0 2 Scsi Port3 Bus0 Disk2 Part0 OPEN NORMAL 40 0 3 Scsi Port3 Bus0 Disk2 Part0 OPEN NORMAL 32 0 DEV 1 DEVICE NAME Disk3 PartO TYPE 2107900 POLICY OPTIMIZED SERIAL 75065711001 LUN IDENTIFIER 6005076303FFC0B60000000000001001 Path Adapter Hard Disk State Mode Select Errors 0 Scsi Port2 Bus0 Disk3 Part0 OPEN NORMAL 6 0 1 Scsi Port2 Bus0 Disk3 Part0 OPEN NORMAL 4 0 2 Scsi Port3 BusO Disk3 Part0 OPEN NORMAL 4 0 3 Scsi Port3 Bus0 Disk3 Part0 OPEN NORMAL 2 0 To obtain the WWPN of your Fibre Channel adapter use the datapath query wwpn command as shown in Example 3 2 Example 3 2 The datapath query command C Program Files IBM Subsystem Device Driver gt datapath query wwpn Adapter Name PortWWN Scsi Port2 210000E08B037575 Scsi Port3 210000E08B033D76 The datapath query essmap and datapath query portmap commands are not available 3 2 2 Mapping SDD devices to Windows drive letters When assigning DS8000 LUNs to a Windows host you must understand which Windows drive letter corresponds to which DS8000 LUN To determine which Windows drive correspond
386. tart adapter Virtual Storage Optical Console adapter Virtual Ethernet Frish cancel ep Figure 11 30 Load source with IVM 11 8 Installing IBM i with boot from SAN through VIOS NPIV This section provides information about installing IBM i on an LPAR of the POWER server to the DS8000 that is to boot from SAN and connected through VIOS NPIV 11 8 1 Scenario configuration The best practice is to connect IBM i with two VIOS NPIV in multipath However for the purpose of the scenario in this section IBM i is connected through two virtual Fibre Channel adapters each of them assigned to a port from a separate physical adapter in the same VIOS 238 IBM System Storage DS8000 Host Attachment and Interoperability Figure 11 31 shows that in IBM i the virtual Fibre Channel adapter ID 14 is connected to the Fibre Channel adapter ID 34 in VIOS and the virtual adapter 15 is connected to the VIOS adapter 35 FES i eyA Ly https 9 155 57 94 hmc wd T233F General Processors Memory Virtual Power Adapters Controlling Logical Partition Profile Properties default_npiv ATS_LPAR19_ITSO_JANA_IP42 p6 570 2 O06C6DE1 ATS_LPAR19_ITSO_JANA_IP42 Logical Host Ethernet Tagged 1 0 Settings Adapters LHEA Actions settings are listed below Maximum virtual adapters Number of virtual adapters Virtual resources allow for the sharing of physical hardware between logic
387. tart scripts to the SLES 11 system start process as shown in Example 6 23 Example 6 23 Configuring automatic start of DM Multipath in SLES 11 x36501ab9 insserv boot multipath x36501ab9 insserv multipathd Enabling multipathing for Red Hat Enterprise Linux 5 Red Hat Enterprise Linux comes with the etc multipath conf default file This file contains a section that blacklists all device types To start DM Multipath remove or comment out these lines A pound sign in front of the comment lines will mark them as comments so they will be ignored the next time DM Multipath scans for devices as shown in Example 6 24 Example 6 24 Disabling blacklisting all devices in etc multiparh conf Blacklist all devices by default Remove this to enable multipathing on the default devices blacklist devnode Start DM Multipath as shown in Example 6 25 Example 6 25 Starting DM Multipath in Red Hat Enterprise Linux 5 root x36501ab9 etc init d multipathd start Starting multipathd daemon OK In order to have DM Multipath start automatically at each system start add the start script to the Red Hat Enterprise Linux 5 system start process as shown in Example 6 26 Example 6 26 Configuring automatic start of DM Multipath in Red Hat Enterprise Linux 5 root x36501ab9 chkconfig add multipathd root x36501ab9 chkconfig levels 35 multipathd on root x36501ab9 chkconfig list multipathd mult
388. tch n 1 multipathd k show top Adding and removing DS8000 volumes in Linux on System z The mechanisms to scan and attach new volumes provided in Adding and removing DS8000 volumes dynamically on Linux on page 106 do not work in the same manner for Linux on System z Commands are available that discover and show the devices connected to the FC HBAs but they do not process the logical attachment to the operating system automatically In SLES10 SP3 use the zfcp_san_disc command for discovery Example 6 39 shows how to discover and list the connected volumes for one remote port or path using the zfcp_san_disc command Run this command for all available remote ports Example 6 39 List LUNs connected through a specific remote port Inxmnt01 zfcp_san_disc L p 0x500507630a00029f b 0 0 0500 0x4048400000000000 0x4049400000000000 0x4049400100000000 0x4049400200000000 Chapter 6 Linux considerations 107 Tip In recent Linux distributions the zfcp_san_disc command is not available Remote ports are automatically discovered The attached volumes can be listed using the 1sluns script After discovering the connected volumes do the logical attachment using sysfs interfaces Remote ports or device paths are represented in sysfs There is a directory for each local or remote port combination or path This directory contains a representative of each attached volume and various meta files as interfaces for action Example 6 40 sho
389. te click OK in the action window shown in Figure 11 24 The action you requested has comp EJ 3 The action you requested has completed successfully Figure 11 24 Disks added successfully to disk pool Chapter 11 IBM i considerations 233 15 Click OK if you see the message that the created IASP is ready to start mirroring The System i Navigator Disk Pools window Figure 11 25 shows the new disk pool Environment My Connections My Connections us Basic Operations A Name Sign Release Description E Work Management fil 9 155 57 42 v7rimO ITSOi I Configuration and Service fil 9 155 70 14 v rimO Mainz I6PFE System Values fil 9 155 70 15 vSr4m0 AS400PFE a ar a fl 9 155 70 32 v rim0 iSPFE a iiini il 9 155 70 40 v7rim0 Manage this system Hardware 4 il Mcei5i1 mai v5r4m0 Manage this system 74 Communications LAN Resources amp Workstation Resources Processor Information E i Cryptography Resources g Optical Units E Disk Units All Disk Units Ej ii By Location Disk Arols Disk Pos Groups E Parity Sets Nonconfigured Disk Units ra Sb Tana Newiree ry Add a connection ha System Adapters Add a connection il Change all of your system passwords Install plug ins Figure 11 25 New disk pool shown in System i Navigator To see the logical volume as shown in Figure 11 26 expand Configuration and Service gt Hardware
390. te a directory for the volume group mkdir dev vg08 b Create a group file within this directory mknod dev vg08 group c 64 0x080000 c Prepare the LUNs being represented by the DSFs to be used in a LVM group pvcreate dev rdsk cl1t0d1 pvcreate dev rdsk cl2t0d1 Chapter 10 HP UX considerations 187 Example 10 7 Volume group creation with earlier DSFs vgcreate A y x y 1 255 p 100 s 16 dev vg08 dev dsk c11t0d1 dev dsk c15t0d1 dev dsk c12t0d1 dev dsk c13t0d1 Volume group dev vg08 has been successfully created Volume Group configuration for dev vg08 has been saved in etc lvmconf vg08 conf vgdisplay v vg08 Volume groups VG Name dev vg08 VG Write Access read write VG Status available Max LV 255 Cur LV 0 Open LV 0 Max PV 100 Cur PV 2 Act PV 2 Max PE per PV 1016 VGDA 4 PE Size MBytes 16 Total PE 1534 Alloc PE 0 Free PE 1534 Total PVG 0 Total Spare PVs 0 Total Spare PVs in use 0 Physical volumes PV Name dev dsk c11t0d1 PV Name dev dsk c15t0d1 Alternate Link PV Status available Total PE 767 Free PE 767 Autoswitch On PV Name dev dsk c12t0d1 PV Name dev dsk c13t0d1 Alternate Link PV Status available Total PE 767 Free PE 767 Autoswitch On In Example 10 7 here the volume group vg08 consists of two LUNs c11t0d1 and c12t0d1 each of which is dual pathed also called Alternate Linked 2 Create the logical volumes in Example 10 7 which are LVM objects that the
391. tecture IBM i only supports specific volume sizes and these volume sizes might not be an exact number of extents In general these volume sizes relate to the volume sizes available with internal devices although certain larger sizes are supported for external storage only IBM i volumes are defined in decimal gigabytes for example 10 bytes When creating the logical volumes for use with IBM i you can see in almost every case that the IBM i device size does not match a whole number of extents and therefore certain space is wasted Table 11 8 provides the IBM i volumes sizes and the number of extents required for IBM i volume sizes Table 11 8 IBM i logical volume sizes Model type IBM i Number Extents Unusable Usable device of logical space space Unprotected Protected size GB block GiB addresses LBAs 2107 A81 2107 A01 8 5 16 777 216 8 0 00 100 00 2107 A82 2107 A02 17 5 34 275 328 17 0 66 96 14 2107 A85 2107 A05 35 1 68 681 728 33 0 25 99 24 2107 A84 2107 A04 70 5 137 822 208 66 0 28 99 57 2107 A86 2107 A06 141 1 275 644 416 132 0 56 99 57 2107 A87 2107 A07 282 2 551 288 832 263 0 13 99 95 1 GiB represents binary gigabytes 28 bytes and GB represents decimal gigabytes 10 bytes Tip IBM i levels 5 4 and 6 1 do not support logical volumes of size 8 59 and 282 2 as an IBM i load source unit or boot disk where the load source unit is located in the extern
392. tem Storage DS8000 Host Attachment and Interoperability 9 6 3 Supplying the VID or PID string In order to take symmetric arrays under control of MPxIO as shown in Example 9 22 on page 174 its VID or PID string must be supplied in the kernel drv scsi_vhci conf file For more information see 9 4 2 Using MPxIO on page 164 However because the operating system used for these examples was installed from scratch there was no possibility to supply the VID or PID string shown in Example 9 23 The DS8000 LUNs are already under control of MPxlO For example DS8000 LUN has one entry instead of two as shown in Example 9 22 on page 174 Example 9 23 here shows the content of the scsi_vhci conf file Example 9 23 Content of the scsi_vhci conf file bash 3 00 more kernel drv scsi_vhci conf Copyright 2004 Sun Microsystems Inc All rights reserved Use is subject to license terms pragma ident scsi_vhci conf 1 9 04 08 26 SMI name scsi_vhci class root Load balancing global configuration setting load balance none will cause all I O to a given device which supports multipath I 0 to occur via one path Setting load balance round robin will cause each path to the device to be used in turn load balance round robin Automatic failback configuration possible values are auto failback enable or auto failback disable auto failback enable For enabling MPxIO support for
393. the snapshot Figure 7 30 Advanced disk options 7 Verify all options in the Ready to Complete window and click Finish Figure 7 31 Options Hardware type Hard Disk Create disk New virtual disk Disk capacity 10 GB Disk provisioning Thick Provision Lazy Zeroed Datastore Local_Disk Virtual Device Node SCSI 0 1 Disk mode Persistent Figure 7 31 Ready to Complete 8 Close the Virtual Machine Properties window 7 2 3 Raw device mappings RDMs Raw device mappings RDMs allow a VM machine to store its data directly on a logical volume instead of writing to a virtual disk file A mapping file stored on the VMFS contains all information the VM requires to access the logical volume Tip RDMs can have a maximum size of 64 TB RDMs allow for SAN aware applications on the VM and provide the possibility to use N Port ID Virtualization NPIV on a virtual machine level See Using NPIV with RDMs on page 141 for details on NPIV 138 IBM System Storage DS8000 Host Attachment and Interoperability Assigning an RDM to a VM The following steps describe how to assign a physical LUN to a VM 1 Select the VM from the list right click and select Edit Settings 2 Click Add 3 From the list of devices select Hard Disk 4 Select Raw Device Mappings as shown in Figure 7 32 A virtual disk is composed of one or more files on the host file system Together these files appear as a single
394. the BIOS for both adapters at the same time If the BIOS for both adapters are enabled at the same time and there is a path failure on the primary adapter the system stops with an INACCESSIBLE BOOT DEVICE error upon reboot 3 9 Windows Server 2003 Virtual Disk Service support With Windows Server 2003 Microsoft introduced the virtual disk service VDS It unifies storage management and provides a single interface for managing block storage virtualization This interface is vendor and technology transparent and is independent of the layer of the operating system software RAID storage hardware and other storage virtualization engines where virtualization is done VDS is a set of APIs that uses two sets of providers to manage storage devices The built in VDS software providers enable you to manage disks and volumes at the operating system level VDS hardware providers supplied by the hardware vendor enable you to manage hardware RAID arrays Windows Server 2003 and 2008 components that work with VDS include the Disk Management Microsoft Management Console MMC snap in the DiskPart command line tool and the DiskKRAID command line tool which is available in the Windows Server 2003 and 2008 deployment kit 32 IBM System Storage DS8000 Host Attachment and Interoperability Figure 3 11 shows the VDS architecture Command Line Tools a eDiskPart Disk Management Storage Management DiskRAID MMC Snap in Applications Software Providers
395. the VIOS partition and AIX client partition In a dual VIOS setup where the same DS8000 volume is assigned to both VIOS partitions and mapped as backing device to the same client partition you must set the reserve_policy attribute on the device to no_reserve which is already the default for DS8000 devices with SDDPCM installed Also the preferred AIX MPIO device attributes algorithm load_balance hcheck_mode nonactive and hcheck_interval 60 seconds are already set by default for DS8000 devices with current SDDPCM releases For dual VIOS setups change the default values of the attributes fc_err_recov and dyntrk of the Fibre Channel adapter s child device fscsiX to the settings shown in Example 4 11 on both virtual I O servers Example 4 11 Changing the fscsi attributes in a dual VIOS setup chdev dev fscsi0 attr fc_err_recov fast_fail dyntrk yes perm fscsi0 changed Changing the fc_err_recov attribute to fast_fail will fail any I Os immediately if the adapter detects a link event such as a lost link between a storage device and a switch The fast_fail setting is only advisable for dual VIOS configurations Setting the dyntrk attribute to yes enables the VIOS to allow cabling changes in the SAN Both virtual I O servers need to be rebooted for these changed attributes to take effect Applying changes Changes applied with the chdev command and the perm option require a restart to become effective Chapter 4 Virtual I O Serv
396. the other nodes in the MSCS To expand a volume while it is in use on Windows 2003 and Windows 2008 use either DiskPart for basic disks or Disk Manager for dynamic disks The DiskPart tool is part of Windows 2003 For other Windows versions you can download it at no cost from Microsoft DiskPart is a tool developed to ease administration of storage It is a command line interface CLI where you can manage disks partitions and volumes using scripts or direct input on the command line You can list disks and volumes select them and after selecting them obtain more detailed information create partitions extend volumes and more For more information see the following Microsoft websites gt Microsoft home page http www microsoft com gt Microsoft support page http support microsoft com default aspx scid kb en us 304736 amp sd tech To list the volume size use the 1sfbvol command as shown in Example 3 7 Example 3 7 Isfbvol fullid before volume expansion dscli gt 1sfbvol fullid 4703 Name ID accstate datastate configstate deviceMIM datatype extpool cap 2 30B cap 10 9B cap blocks ITSO_x346_3_4703 IBM 2107 7520781 4703 Online Normal Normal 2107 900 FB 512 IBM 2107 7520781 P53 12 0 25165824 In Example 3 7 here you can see that the capacity is 12 GB and also what the volume ID is Chapter 3 Windows considerations 27 To discover what disk this volume is on in the Windows 2003 host use the datapath query dev
397. the specified name already exists use the f option to force mkinitrd to overwrite the existing one For a more detailed output use the v option 92 IBM System Storage DS8000 Host Attachment and Interoperability Determine the kernel version that is currently running on the system with the uname command as shown in Example 6 10 Example 6 10 Determining the Kernel version root x36501ab9 uname r 2 6 18 194 e15 6 2 3 Determining the WWPN of installed HBAs To create a host connection on the DS8000 that provides the ability to map volumes to a certain HBA you need the HBA s WWPN The WWPN along with additional information about the HBA is provided in the sysfs file The sysfs file is a Linux pseudo file system that reflects the installed hardware and its configuration Example 6 11 shows how to determine which SCSI host instances are assigned to the installed FC HBAs and their WWPNs Example 6 11 Determining the WWPNs of the FC HBAs root x36501ab9 1s sys class fc_host host3 host4 root x36501ab9 cat sys class fc_host host3 port_name 0x2100001b32095bad root x36501ab9 cat sys class fc_host host4 port_name 0x2100001b320969b6 Tip The sysfs file contains additional information and is used to modify the hardware configuration Additional information is provided for sysfs in Example 6 40 on page 108 6 2 4 Checking attached volumes The easiest method to check for DS8000 volumes attached to the Linux
398. them to appear in IBM i At this stage they are used in exactly the same method as for non configured internal units There is nothing particular to external logical volumes as far as IBM i is concerned You use the same functions for adding the logical units to an ASP as you use for internal disks When connecting DS8000 natively or with VIOS NPIV the volumes are shown in IBM i as disk units with the following characteristics gt Type 2107 gt Model AOx or A8x The model depends on the size and protection gt Serial number The serial number is in the form 50 yyyyzzz Where yyyy stands for the LUN ID and zzz stands for the last three characters of the DS8000 WWPN gt Resource name The resource name starts with DMP when using multipath and with DD when using single path 224 IBM System Storage DS8000 Host Attachment and Interoperability When connecting DS8000 to IBM i with VIOS VSCSI the volumes appear in IBM i with the following attributes gt Type 6B22 gt Model 050 gt Serial number with 12 characters You can add volumes to IBM i configuration by using the IBM i 5250 interface Operations Navigator GUI or the web GUI Systems Director Navigator for IBM i These methods can be used for adding LUNs to IBM i System ASP user ASPs or IASPs Tip The web GUI Systems Director Navigator for i is available with IBM i V6R1 and later 11 6 9 Using the 5250 interface You can add disk units to the configuration eith
399. thing Each SAN Volume Controller node presents a VDisk to the SAN via multiple paths A typical configuration is to have four paths per VDisk In normal operation two nodes provide redundant paths to the same storage which means that depending on zoning and SAN architecture a single server might see eight paths to each LUN presented by the SAN Volume Controller Each server host bus adapter HBA port needs to be zoned to a single port on each SAN Volume Controller node Because most operating systems cannot resolve multiple paths back to a single physical device IBM provides a multipathing device driver The multipathing driver supported by the SAN Volume Controller is the IBM Subsystem Device Driver SDD SDD groups all available paths to a virtual disk device and presents it to the operating system SDD performs all the path handling and selects the active I O paths SDD supports the concurrent attachment of various DS8000 models Storwize V7000 and SAN Volume Controller storage systems to the same host system Where one or more alternate storage systems are to be attached you can identify the required version of SDD at this website http www ibm com support docview wss uid ssg1 7001350 13 3 Configuration guidelines for SVC The following sections give general guidelines for attaching the SVC to DS8000 13 3 1 Determining the number of controller ports for DS8000 To avoid performance issues it is important to configure a min
400. this attribute is 120 seconds and it is user changeable with the valid range of 30 to 600 seconds The pempath command provides a new CLI command to dynamically change this device retry_timeout attribute Here is the syntax of this command pcmpath set device lt device number gt retry timeout lt time gt When only num is specified for the device number then the command applies to the hdisk specified by that num Chapter 5 AIX considerations 75 This feature enables the user to adjust the time out value to control how long SDDPCM will retry the I O on the last path before it fails to the application In the situation where a device loss of access is only temporary the retry_timeout value might need to be set to a higher value SDDPMC server daemon The SDDPCM server also referred to as pemsrv is an integrated component of SDDPCM 2 1 0 0 or later This component consists of a UNIX application daemon that is installed in addition to the SDDPCM path control module The SDDPCM server daemon provides a path recovery function for SDDPCM devices and the First Time Data Capture function After you have installed SDDPCM and restarted the system verify if the SDDPCM server pemsrv has automatically started as shown in Example 5 20 Example 5 20 verifying if SDDPCM has automatically started Issrc s pcmsrv Subsystem Group PID Status pcmsrv 4718636 active It is possible to start the SDDPCM server manually as follows gt For SDD
401. tically start with managed system Disabled Service and support Connection monitoring Disabled Service partition Disabled Redundant error path reporting Disabled Electronically report errors that cause partition termination or require attention Time reference Disabled Disabled Tagged I O Load source Slot 14 Virtual 1 0 Alternate restart Slot 2 Virtual 1 0 Console Slot 1 Virtual I O Alternate console Operations console direct OK Cancel Help one Figure 11 36 IPL settings To start the installation activate the IBM i LPAR and then complete the following steps 1 Select the language group for IBM i and then select option 1 Install Licensed Internal Code Figure 11 37 Install Licensed Internal Code System TOOC6DE1 Select one of the following 1 Install Licensed Internal Code 2 Work with Dedicated Service Tools DST 3 Define alternate installation device Selection 1 Figure 11 37 Installing License Internal Code 242 IBM System Storage DS8000 Host Attachment and Interoperability 2 Select the LUN to be the load source device in the Select Load Source Device window as shown in Figure 11 38 1 50 210A29F 2107 50 210929F 2107 50 200929F 2107 50 200829F 2107 F3 Exit F5 Refresh Type 1 to select press Enter Opt Serial Number Type Model A04 A04 A04 A04 Sys Bus 255 255 255 255 Sys Card 14 15 14 15 Select Load Source
402. ting using SAN switches 000 00 cece ee 217 TES Using multipath resa i wkd ie Re we eke Ghee ah da ea ee 217 11 5 1 Avoiding single points of failure 00 ee 218 11 5 2 Configuring the multipath 0 2 0 0 eee 218 11 5 3 Multipathing rules for multiple System i hosts or partitions 219 11 6 Configuration guidelines 0 0 00 tenes 220 11 6 1 Creating extent pools for IBMiLUNS 0 000 c eee eee 220 11 6 2 Defining the LUNs for IBMi 2 0 2 eee 220 11 6 3 Protected versus unprotected volumes 0 00 eee eee eee 221 11 6 4 Changing LUN protection 0 0 00 c cee 222 11 6 5 Setting the ports and defining host connections for IBMi 222 11 6 6 Zoning the switches 0 0 sn cece eae 223 11 6 7 Setting up VIOS 0 0 teas 223 11 6 8 Adding volumes to the System i configuration 00 00000 224 11 6 9 Using the 5250 interface 0 eee 225 11 6 10 Adding volumes to an independent auxiliary storage pool 228 11 7 Booting from SAN aisean s auii ahlindees Dacca E AE Ake es 235 11 7 1 Requirements for boot from SAN 0 000 cee 235 11 7 2 Tagging the load source LUN 2 0 ee eee 235 11 8 Installing IBM i with boot from SAN through VIOS NPIV 200000 238 11 8 1 Scenario configuration 0 0 0 eae 238 11 8 2 Booting from SAN and cloning 20 0c eee 243 TAOS MIQKAtING enora A atria Sake
403. tion policy Inter physical volume allocation is one of the simplest and most advisable methods to spread the workload accesses across physical resources Most LVMs offer inter disk allocation policies for logical volumes This method can also be recognized when the term physical partition spreading is used With AIX LVM one or more volume groups can be created using the physical disks which are logical volumes on the DS8000 LVM organizes volume group space in physical partitions Use physical partition allocation for the logical volumes in rotating order known as round robin With this method the first free extent is allocated from the first available physical volume The next free extent is allocated from the next available physical volume and so on If the physical volumes have the same size optimal I O load distribution among the available physical volumes can be achieved 5 5 3 LVM mirroring LVM has the capability to mirror logical volumes across several physical disks This improves availability because when a disk fails there is another disk with the same data When creating mirrored copies of logical volumes ensure that the copies are distributed across separate disks With the introduction of SAN technology LVM mirroring can help provide protection against a site failure Using longwave Fibre Channel connections a mirror can be stretched up to a 10 km 6 2 mi distance 5 5 4 Impact of DS8000 storage pool striping Starti
404. tions 0 0 00 cee eee 201 11 1 Supported environment sssaaa anaana aeaea 202 11 1 1 System hardware 0 0 cette eee 202 TAD Software psie siaa eigen he whew a we Reais Gad Ge ieee a eels ato SE 202 11 1 3 Overview of hardware and software requirements 2005 203 Titae Useful Websites aa oa ie wee ee eo ake EE SR ea eases bed a aes 204 11 2 Using Fibre Channel adapters 0000 c cee 205 11 2 1 Native attachment perius ererken ee eae 205 11 2 2 Attachment with VIOS NPIV 0 2 0 0 000 eee 206 11 2 3 Attachment with VIOS VSCSI 00 0 00 cee 206 11 2 4 Overview of the number of LUNs per adapter and the queue depth 207 11 3 Sizing and implementation guidelines 0 000 cee ee 207 11 3 1 Sizing for natively connected and VIOS connected DS8000 208 11 3 2 Planning for arrays and DDMS 0 0 00 eee eee 208 AALS GAC sei ace ee Abe tae Ne asta satel A ell ne ae vee I 208 11 3 4 Number of ranks 0 0 ee eee 209 11 3 5 Sizing for SSD and hot data management with IBMi 210 11 3 6 Number of Fibre Channel adapters in IBM i and in VIOS 213 11 4 Sizing and numbering of LUNS 000000 eee 215 IBM System Storage DS8000 Host Attachment and Interoperability 11 4 1 Logical volume sizes e ranei kerana a ccc es 215 11 4 2 Sharing or dedicating ranks for an IBM i workload 2 05 216 11 4 3 Connec
405. tions function creates a bit mask that indicates which pages contain data and must be written Only the pages that are populated with data are written to the storage system IBM i native attaches to the DS8000 supports skip operations However VIOS requires that they are repeated by breaking up the non contiguous I O block into multiple discrete contiguous I O blocks Consequently you see more writes per second and smaller transfer sizes in the environment with VIOS as compared to native attachment There is no difference in the overall workload throughput in MBps whether the DS8000 is connected natively to IBM i or through VIOS This point is valid for both VIOS NPIV and VIOS VSCSI Therefore our advice is to use the same sizing guidelines for both natively and VIOS connected DS8000 Sizing Although the characteristics of the IBM i workload slightly change when attaching the DS8000 through VIOS use the same sizing guidelines that apply for connecting natively to IBM i 11 3 2 Planning for arrays and DDMs For IBM i production workloads use 146 GB 15 000 revolutions per minute RPM disk drive modules DDMs or 300 GB 15 000 RPM DDMs The larger slower drives might be suitable for less I O intensive work or for those workloads that do not require critical response times for example archived data or data that is high in volume but low in use such as scanned images Typical IBM i workloads will benefit from using solid state drives
406. tipath tool scans the device or path configuration and builds the instructions for the device mapper These include the composition of the multipath devices failover failback patterns and load balancing behavior Future functionality of the tool will be moved to the multipath background daemon and will not be available in future releases gt The multipath background daemon multipathd constantly monitors the state of the multipath devices and the paths In case of events it triggers failover and failback activities in the DM Multipath module It also provides a user interface for online reconfiguration of the multipathing In the future it will take over all configuration and setup tasks gt A set of rules that tell udev what device nodes to create so that multipath devices can be accessed and are persistent Configuring DM Multipath Use the etc multipath conf file to configure DM Multipath to gt Define new storage device types Exclude certain devices or device types Set names for multipath devices Change error recovery behavior Yv y For more information about etc multipath conf see 6 1 2 Attaching Linux server to DS8000 resources on page 82 In 6 2 9 Attaching DS8000 Considerations on page 104 also provides settings that are preferred specifically for DS8000 attachment You can configure DM Multipath to generate user friendly device names by specifying this option in etc multipath conf as shown in Example 6
407. tive VIOS is connected only with one SAN and a switch in that SAN fails IBM i misses the LUNs However if each VIOS is connected to both SANs resiliency occurs Figure 11 4 shows an example of connecting two VIOS through two SANs POWER VIOS 1 IBM i VIOS 2 FC FC Adapters Virtual FC Adapters i adapters l SAN A SAN B HBAs DS8000 HBAs Figure 11 4 Connecting with two VIOS and two SANs VIOS VSCSI assigns the volumes to two ore more Fibre Channel adapters each of them in a separate VIOS The volumes in VIOS or hdisks can then be mapped to the separate virtual SCSI adapters in IBM i As specified in 11 6 7 Setting up VIOS on page 223 change the SCSI reservation attribute for the LUNs in each VIOS to non reserve After the LUNs are assigned to IBM i IBM i recognizes them as multipathed LUNs and establishes two or more paths to them To achieve multipathing you do not need to set up any driver in IBM i or perform any further tasks Native attachment multipathing With native attachment to the DS8000 multipathing can be established for the Load Source LUN Boot LUN when connecting with IOP based or IOP less adapters on IBM i levels V6R1 or V7R1 However multipathing is not possible for the Load Source LUN on IBM i V5R4 11 5 3 Multipathing rules for multiple System i hosts or partitions When you use multipath disk units consider the implications of
408. torage Adapters view Details vmhba5 Model ISP2532 based 8Gb Fibre Channel to PCI Express HBA WWN 20 00 00 24 ff 2d c5 69 21 00 00 24 ff 2d c5 69 Targets 2 Devices 4 Paths 8 View Devices Paths Name Runtime Name LUN Type Drive Type Transport Capacity Owner IBM Fibre Channel Disk naa 6005076 wmbhba2 C0 T0 LO 0 disk Non SSD Fibre Chann 5 00 GB NMP IBM Fibre Channel Disk naa 6005076 wmbhba2 CO TO L1 1 disk Non SSD Fibre Chann 40 00 GB NMP IBM Fibre Channel Disk naa 6005076 wmhba2 C0 T0 L2 2 disk _Naon SSD Fibre Chann 25 00 GB NMP IBM Fibre Channel Disk naa 6005076 wmhba2 C0 T0 L3 3 WO Fibre Chann 100 00 GB NMP Figure 7 47 SSD detection To use such a volume as host cache first a VMFS datastore must be created on the volume Creating a VMFS datastore on page 126 describes the required steps In the Configuration tab select Host Cache Configuration as shown in Figure 7 48 Host Cache Configuration SSD Datastores Refresh Add Storage RescanAll Properties Identification Capacity Host Cache Space Free Space B DS8000_SDD_DS1 99 75 GB 0 00 B 98 80 GB Figure 7 48 SSD datastore list Select the SSD datastore from the list and click Properties Chapter 7 VMware vSphere considerations 147 Set the check mark at Allocate space for host cache and choose how much space to use as cache as shown in Figure 7 49 a Host Cache Configuration DS8000_SDD_DS1i
409. ttachment and Interoperability Virtual I O Server considerations This chapter provides an overview of attaching IBM Virtual I O Server VIOS on IBM Power Systems to DS8000 storage systems The following topics are covered gt Working with IBM Virtual I O Server gt gt Using VSCSI with IBM VIOS and DS8000 Using NPIV with IBM VIOS and DS8000 For more information about IBM PowerVM virtualization and the management of virtual storage devices on IBM Power Systems with VIOS see the following documentation gt IBM PowerVM Getting Started Guide REDP 4815 http www redbooks ibm com abstracts redp4815 html 0pen IBM PowerVM Live Partition Mobility SG24 7460 http www redbooks ibm com abstracts sg247460 html 0pen IBM PowerVM Virtualization Managing and Monitoring SG24 7590 http www redbooks ibm com abstracts sg247590 html 0pen PowerVM and SAN Copy Services REDP 4610 http www redbooks ibm com abstracts redp4610 html 0pen PowerVM Migration from Physical to Virtual Storage SG24 7825 http www redbooks ibm com abstracts sg247825 html 0pen PowerVM Virtualization on IBM System p Introduction and Configuration Fourth Edition SG24 7940 http www redbooks ibm com abstracts sg247940 html 0pen Copyright IBM Corp 2011 2013 All rights reserved 47 4 1 Working with IBM Virtual I O Server Virtualization is a key factor of IBM Smarter Planet IBM Dynamic Infrastructure strategy On IBM Power
410. ty of MPIO is limited but provides an interface for vendor specific PCMs that allow for implementation of advanced algorithms IBM provides a PCM for DS8000 and other devices that enhances MPIO with all the features of the original SDD Benefits of MPIO There are several benefits of MPIO with SDDPCM compared to traditional SDD gt gt gt Performance improvements due to direct integration with AIX Better integration if separate storage systems are attached Easier administration Requirements and considerations The following requirements and considerations apply for MPIO gt gt gt Default MPIO is not supported on DS8000 SDDPCM is required Both SDDPCM and SDD cannot be installed on the same AIX server If you use MPIO with SDDPCM instead of SDD remove the regular DS8000 host attachment script and install the MPIO version The MPIO version identifies the DS8000 volumes to the operating system as MPIO manageable 74 IBM System Storage DS8000 Host Attachment and Interoperability In comparison to SDD where each path is represented by hdisk and a special vpath device is created SDDPCM does not create a special device for every path See Example 5 18 Example 5 18 The pcmpath query device pcmpath query device 12 13 DEV 12 DEVICE NAME hdisk12 TYPE 2107900 ALGORITHM Load Balance SERIAL 75TV1816200 Path Adapter Path Name State Mode Select Errors 0 fscsi0 path0 CLOSE NORMAL 0 0 1 fscsi0 path1 CLOSE
411. ual Fibre Channel and NPIV see IBM PowerVM Virtualization Managing and Monitoring SG24 7590 LUNs IBM i supports DS8000 LUNs attached to VIOS using NPIV IBM i multipath is automatically started on the DS8000 LUNs that are connected through two VIOS partitions in NPIV Up to 64 LUNs can be assigned to a virtual port in IBM i in NPIV the queue depth is 6 operations per LUN On IBM Power Systems with PowerVM a pair of WWPNs is generated for a virtual Fibre Channel client adapter with subsequent WWPNs as shown in Figure 5 1 on page 64 Only one WWPN is active on a managed system at a time The second WWPN is used for the PowerVM Live Partition Mobility feature and becomes active on the secondary managed system when a partition with virtual Fibre Channel adapters is migrated to this managed system In order to maintain access to the mapped volumes on DS8000 storage systems with PowerVM Live Partition Mobility create additional host connections and SAN zoning also for these secondary WWPNs For more information about IBM PowerVM Live Partition Mobility see BM PowerVM Live Partition Mobility SG24 7460 IBM System Storage DS8000 Host Attachment and Interoperability Copyright IBM Corp 2011 2013 All rights reserved AIX considerations This chapter provides information about the IBM AIX operating system It is not intended to repeat general information that is in other publications For general information see the IBM System Storage
412. ult value is NO and prevents allocating data sets to an EAV volume Example 12 5 shows a sample message that you receive when trying to allocate a data set on EAV volume and USEEAV NO is set Example 12 5 Message IEF021 with USEEVA set to NO IEF021I TEAM142 STEP1 DD1 EXTENDED ADDRESS VOLUME USE PREVENTED DUE TO SMS USEEAV NO SPECIFICATION Chapter 12 IBM System z considerations 259 A new parameter called break point value BPV determines the size of a data set that can be allocated on a cylinder managed area The default for the parameter is 10 cylinders and it can be set on PARMLIB member IGDSMSxx and in the storage group definition Storage group BPV overrides system level BPV The BPV value can be 0 65520 The 0 value means that the cylinder managed area is always preferred and the 65520 value means that a track managed area is always preferred Prerequisite Before implementing EAV volumes apply the latest maintenance z OS V1 10 and higher levels coexisting maintenance levels Space considerations for data set allocation For data set allocation and space requirements keep in mind the following considerations gt When allocating a data set extent on an EAV volume the space requested is rounded up to the next MCU if the entire extent is allocated in cylinder managed space Individual extents always start and end on an MCU boundary gt A given extent is contained in a single managed space which means an extent cannot
413. ults obtained in other operating environments may vary significantly Some measurements may have been made on development level systems and there is no guarantee that these measurements will be the same on generally available systems Furthermore some measurements may have been estimated through extrapolation Actual results may vary Users of this document should verify the applicable data for their specific environment Information concerning non IBM products was obtained from the suppliers of those products their published announcements or other publicly available sources IBM has not tested those products and cannot confirm the accuracy of performance compatibility or any other claims related to non IBM products Questions on the capabilities of non IBM products should be addressed to the suppliers of those products This information contains examples of data and reports used in daily business operations To illustrate them as completely as possible the examples include the names of individuals companies brands and products All of these names are fictitious and any similarity to the names and addresses used by an actual business enterprise is entirely coincidental COPYRIGHT LICENSE This information contains sample application programs in source language which illustrate programming techniques on various operating platforms You may copy modify and distribute these sample programs in any form without payment to IBM for the purposes of
414. umes gt Track managed space is the area on an EAV located within the first 65 520 cylinders Using the 16 bit cylinder addressing enables a theoretical maximum address of 65 535 cylinders In order to allocate more cylinders you need to have a new format to address the area of more than 65 520 cylinders With 16 bit cylinder numbers the existing track address format is CCCCHHHH The 16 bit track number is represented with HHHH The 16 bit track cylinder is represented with CCCC gt Cylinder managed space is the area on an EAV located after the first 65 520 cylinders This space is allocated in multicylinder units MCU which have a size of 21cylinders To address the extended capacity on an EAV the new cylinder track address format for 28 bit cylinder numbers is CCCCcccH The H represents a 4 bit track number from zero to 14 The high order 12 bits of a 28 bit cylinder number is represented with ccc The low order 16 bits of a 28 bit cylinder number is represented with CCCC Components and products such as DADSM CVAF DFSMSdss ICKDSF and DFSORT also now support 1 182 006 cylinders Consider the following examples gt DS8000 and z OS limit CKD EAV volume size For 3390 Model A one to 1 182 006 cylinders or about 1 TB of addressable storage For 3390 Model A up to 1062 times the size of 3390 Model 1 gt Configuration granularity For 1 cylinder boundary sizes 1 to 56 520 cylinders For 11
415. ust be presented as a physical device hdiskX from the Virtual I O Servers to the client partition It is not possible to provide a single DS8000 LUN and then further subdivide it into logical volumes as backing devices at the VIOS level when intending to use a redundant setup with two VIOS and multipathing in the client partition Chapter 4 Virtual I O Server considerations 53 Using SDDPCM for example the SDD path control module for AIX MPIO on the VIOS the DS8000 LUNs is displayed as IBM MPIO FC 2107 devices when properly configured as shown in Example 4 5 Example 4 5 VIOS with attached DS8000 volumes using SDDPCM 1sdev type disk name status description hdisk0O Available IBM MPIO FC 2107 hdiskl Available IBM MPIO FC 2107 hdisk2 Available IBM MPIO FC 2107 The properties of the DS8000 LUNs using SDDPCM 2 6 0 0 are shown in Example 4 6 You can get the serial number of the DS8000 storage system and the DS8000 volume ID from the output of the Isdev dev hdiskX vpd command The Serial Number 75TV1819000 displayed includes the serial number 75TV181 of the storage image on the DS8000 storage system and the DS8000 volume ID 9000 Example 4 6 Serial number of DS8000 LUNs in Virtual I O Server Isdev dev hdiskO vpd hdisk0 U78C0 001 DBJ2743 P2 C4 T1 W500507630A4B429F L4090400000000000 IBM MPIO FC 2107 Manufacturer sesssesesese IBM Machine Type and Model 2107900 Serial Number 75TV1819000 EC k ye lrristen
416. uture WWPNs for virtual Fibre Channel adapters careful attention is required to manage partition profiles with virtual Fibre Channel adapters to maintain WWPNs and not lose the work done for the device mapping on the storage system and SAN zoning Continuous availability is achieved with a redundant VIO server setup such as the following example gt Using a minimum of four virtual Fibre Channel adapters in the client partition gt Connecting to both VIO server partitions gt Mapping to separate physical Fibre Channel adapters and two SAN fabrics Chapter 4 Virtual I O Server considerations 59 60 A redundant VIO server setup Figure 4 6 provides a reasonable redundancy against a VIOS failure A VIOS failure can be due to a planned maintenance and a simultaneous failure of one SAN fabric Virtual I O server 1 Virtual I O server 2 Client logical partition SAN Client virtual to 7 fibre channel Physical fibre Physical fibre adapter 1 channel adapter channel adapter Client virtual fibre channel Physical fibre Physical fibre adapter 2 channel adapter channel adapter Client virtual Server virtual fibre Server virtual fibre fibre channel channel adapter channel adapter adapter 3 Server virtual fibre Server virtual fibre Fesh basos channel adapter channel adapter adapter 4 Hypervisor Figure 4 6 Virtual Fibre Channel on a dual VIOS managed system For more information about virt
417. ux installer sets the boot device in the firmware to the drive on which you are installing You do not need to take any special precautions regardless of whether you install on a local disk a SAN attached DS8000 volume or a virtual disk provided by the VIOS IBM System z Linux on System z can be loaded from traditional CKD devices or from Fibre Channel attached fixed block SCSI devices To load from SCSI disks the SCSI feature FC 9904 named initial program load IPL must be installed and activated on the System z server SCSI IPL is generally available on recent IBM System z machines IBM z10 and later Attention Activating the SCSI IPL feature is disruptive It requires a power on reset POR of the whole system Linux on System z can run in the following two separate configurations gt Linux on System z running natively in a System z LPAR After installing Linux on System z you need to provide the device from where the LPAR runs the IPL in the LPAR start dialog on the System z support element After it is registered there the IPL device entry is permanent until changed gt Linux on System z running under z VM Within z VM start an operating system with the IPL command With the IPL command you must provide the z VM device address where the Linux boot loader and kernel is installed When booting from SCSI disk the disk does not have a z VM device address For more information see System z on page 88 You must p
418. v hosts v E localhost gt E pci1000 50 Status Online Load Balance Pressure Transfer Count 431 003 MB v E Celerity FC 82EN Channel 1 Paths Channel 2 v MultiPath Director Target Port Read Mode Write Mode Transferred y 50 05 07 63 0a 00 42 9f Preferred Preferred E uno 50 05 07 63 0a 40 42 9f Preferred Preferred 40 05 KB LUN 1 K Target Adapter Statistics Node 20 00 00 10 86 03 29 a8 PCI Location 10 0 0 Port 21 00 00 10 86 03 29 a8 Domain ID 1 Port ID 202F00 Topology Point To Point Link Up Link Speed 4 Gb s rStatus a 2107900 information retrieved a Figure 8 1 ATTO Configuration Tool Paths The path information displayed is based on a per Target or per LUN basis depending on what device is selected in the Device Listing pane IBM System Storage DS8000 Host Attachment and Interoperability 8 4 2 Target Base Click the Target in the device tree 2107900 in Figure 8 1 on page 154 to display the Target Multipathing window with all available paths to the target The icon to the left of the device indicates the multipathing status for the target as shown in Figure 8 2 Path Status at Target Level GC A singe path is connected to your device Multiple paths are connected to your device and are functioning normally Multiple paths were comected to your device but one or more have failed X Miltiple paths were connected to your device but all have failed
419. v sda3 Kernel Modules hwmon thermal sys scsi_transport_fc qla2xxx module qla2xxx ko firmware 1ib firmware q12500_fw bin module qla2xxx ko Features block usb resume userspace resume kernel Bootsplash SLES 800x600 30015 blocks If you need nonstandard settings for example a separate image name use parameters for the mkinitrd command For additional information see the man page for the mkinitrd command on your Linux system Red Hat Enterprise Linux Kernel modules that must be included in the initRAMFS are listed in the file etc modprobe conf The order that they appear in the file as shown in Example 6 8 is the order they will be loaded at system startup Example 6 8 Configuring Red Hat Enterprise Linux to include a kernel module in the initRAMFS root x36501ab9 cat etc modprobe conf alias eth0 bnx2 alias ethl bnx2 alias eth2 e1000e alias eth3 e1000e alias scsi_hostadapter aacraid alias scsi_hostadapterl ata_piix alias scsi_hostadapter2 qla2xxx alias scsi_hostadapter3 usb storage After adding the HBA driver module to the configuration file rebuild the initRAMFS with the mkinitrd command The Red Hat version of mkinitrd requires parameters such as the name and location of the image file to create and the kernel version it is built for as shown in Example 6 9 Example 6 9 Creating the initRAMFS root x36501ab9 mkinitrd boot initrd 2 6 18 194 e15 img 2 6 18 194 e15 If an image file with
420. vices allowing access to the devices and drivers through a virtualization layer One of the types of DSF is the persistent device special file which references a device by a unique name and is not dependent on any specific hardware path With HP UX 11i Version 3 HP introduced several major changes especially in the I O area One of these is an alternate naming scheme for device special files gt Persistent DSFs are created independently from the underlying hardware path information and the number of physical paths to a device The generic syntax of a persistent DSF for a LUN is dev r disk diskN with N being the device instance number This kind of representation scheme is called the agile view for example dev r disk disk14 gt The older format for DSFs such as dev r dsk c12t0d1 is still available and is called legacy DSF in HP UX 11iv3 Accordingly this representation is called the legacy view 10 5 2 Device discovery After you configure the volumes on the DS8000 you can connect your host to the fabric then discover the devices by using the ioscan command Example 10 4 shows that the DS8000 devices were discovered successfully but the devices cannot be used yet because no special device file is available Example 10 4 Discovered DS8000 devices without a special device file ioscan fnC disk Class I H W Path Driver S W State H W Type Description disk 1 0 0 2 1 0 16 UsbScsiAdaptor CLAIMED DEVICE USB SCSI Stack Adaptor
421. vpd command for the particular adapter X 0 1 2 and so on in the VIOS and look for the network address as shown in Example 4 1 In this example all the Fibre Channel adapters are shown with the corresponding WWN Example 4 1 List WWPNs of the Fibre Channel adapters in the Virtual I O Server Isdev dev fcs while read a b do echo a 1sdev dev a vpd grep Network done name fcsO Network Address 10000000C9998F6E fcsl Network Address 10000000C9998F6F 50 IBM System Storage DS8000 Host Attachment and Interoperability To assign volumes to the VIOS on the DS8000 storage system complete the following steps 1 Create the DS8000 volume group for the volumes accessed by the VIOS using the mkvolgrp command with the hosttype IBM pSeries parameter 2 Create DS8000 host connections for the Fibre Channel adapters in VIOS using the mkhostconnect command with the hosttype pSeries parameter and an unused port group for example port group 60 to group multiple host connections of the same host or hosts which share the same volumes together 3 Create and assign DS8000 volumes to the DS8000 volume group using the mkfbvol command Example 4 2 shows the use of extent pools P8 and P9 as storage pools and volume IDs from the logical subsystem LSS 60 rank group 0 and LSS 61 rank group 1 4 Assign the DS8000 volume group to the DS8000 host definition for the VIOS using the managehostconnect command Exa
422. ware_products System_Storage_Multipath_Subsystem Device_Driver After you install the SDD software you can see that the paths were grouped into virtual vpath devices Example 9 5 shows the output of the showvpath command Example 9 5 Output of the showvpath command opt IBMsdd bin showvpath vpath1 Serial Number 75207814206 c2tld1s0 devices pci 6 4000 fibre channel 2 sd 1 1 a raw c3t1d1s0 devices pci 6 2000 fibre channel 1 sd 1 1 a raw vpath2 Serial Number 75207814205 c2t1d0s0 devices pci 6 4000 fibre channel 2 sd 1 0 a raw c3t1d0s0 devices pci 6 2000 fibre channel 1 sd 1 0 a raw For each device the operating system creates a node in the dev dsk and dev rdsk directories After SDD is installed you can see these new vpaths by listing the contents of those directories With SDD the old paths are not suppressed Instead new vpath devices show up as dev rdsk vpathla for example When creating volumes and file systems be sure to use the vpath device instead of the original device Chapter 9 Solaris considerations 163 SDD also offers parameters that you can tune for your environment Specifically SDD offers three separate load balancing schemes gt Failover No load balancing The second path is used only if the preferred path fails gt Round robin The paths used are chosen at random but separate paths than the most recent I O If there are only two paths then they alternate gt Load
423. ween persistent DSFs and earlier DSFs as shown in Example 10 6 Example 10 6 Relationship between persistent DSFs and earlier DSFs ioscan Nm dsf Persistent DSF dev rdisk disk12 dev rdisk disk13 dev rdisk disk14 Legacy DSF s dev rdsk c2t0d0 dev rdsk c3t0d0 dev rdsk c0t0d0 dev rdsk c1t0d0 dev rdsk c2t0d1 dev rdsk c3t0d1 dev rdsk c0t0d1 dev rdsk c1t0d1 dev rdsk c30t0d0 186 IBM System Storage DS8000 Host Attachment and Interoperability dev rdisk disk14 pl dev rdsk c30t0d0s1 dev rdisk disk14_p2 dev rdsk c30t0d0s2 dev rdisk disk14_p3 dev rdsk c30t0d0s3 dev rdisk disk15 dev rdsk c30t1d0 The results of the ioscan Nm command display all of the connection paths for agile devices Example 10 6 on page 186 shows that the agile device dev rdisk disk12 is connected through four paths dev rdsk c2t0d0 and dev rdsk c3t0d0 and dev rdsk cOt0d0 and dev rdsk c1t0d0 When the volumes are visible you can then create volume groups VGs logical volumes and file systems 10 6 Multipathing The IBM SSIC indicates that the DS8000 supports native multipathing with HP UX Along with the operating system delivery HP also ships a version of Symantec s Veritas Volume Manager VxVM which is a software product designed to enable logical management of physical devices However a license for Dynamic Multipathing DMP is not included This implies that any I O is handled in pass through mode and is executed by nativ
424. with a suspended volume presents an unsolicited unit check to each hot spot that is grouped on the LSS indicating PPRC suspension Software initiates a task to issue two commands to get PPRC state of volume on LSS and reset summary event state First volume PPRC suspension event on LSS is reported immediately to all grouped hosts gt If all hosts accept an event the next event notification waits at least 150 ms gt Incase some host does not accept the event in 5 seconds the state change status on each affected volume is presented New storage controller health attention message for new volume events Event code severity code and LSS device bit map in each message IBM GDPS uses message severity code to initiate PPRC failover CKD Aggregation Summary is supported on z OS starting from level V1R13 Figure 12 5 shows an example of the difference of an error that occurred with and without Error Aggregation As you can see the difference is conspicuous not only with the total amount of errors reported but also in how the errors are managed Site 1 Metro Mirror al Sysplex od PPRC X One volume 64K Vols 64K Vol S4LPARS on 255 LSS suspends 255 LSS Without aggregation With aggregation One volume suspends One volume suspends 1 state change to each of 64 LPARs 1 unit check to each of 64 LPARs GDPS freezes 255 LSS GDPS freezes 255 LSS 65519 state changes to each of
425. ws a sysfs structure for a specific DS8000 remote port Example 6 40 sysfs structure for a remote port Inxmnt01 1s 1 sys bus ccw devices 0 0 0500 0x500507630a00029f total 0 drwxr xr x 2 root root 0 2010 11 18 11 03 0x4049400000000000 W 1 root root 0 2010 11 18 17 27 unit_add W 1 root root 0 2010 11 18 17 55 unit_remove In Example 6 41 LUN 0x4049400100000000 is added to all available paths using the unit_add metafile Example 6 41 Adding a volume to all existing remote ports Inxmnt01 echo 0x4049400100000000 gt sys 0 0 0500 0x500507630a00029f unit_add Inxmnt01 echo 0x4049400100000000 gt sys 0 0 0500 0x500507630a0b029f unit_add Inxmnt01 echo 0x4049400100000000 gt sys 0 0 0600 0x500507630a03029f unit_add Inxmnt01 echo 0x4049400100000000 gt sys 0 0 0600 0x500507630a08029f unit_add Important Perform discovery using zfcp_san_disc when new devices such as remote ports or volumes are attached Otherwise the system does not recognize them even if you do the logical configuration New disk devices attached in this method will automatically get device nodes and are added to DM Multipath Removing a volume from Linux on System z To remove a volume from Linux on System z and avoid system suspending due to incomplete I O requests follow the same sequence used in Removing a disk device or volume from Linux on page 107 Volumes can then be removed logically using a met
426. ws the DS8000 I O ports where the Fibre Channel adapters of the AIX server are logged in Example 5 4 Checking DS8000 volume assignment to an AIX server using DSCLI dscli gt Ishostconnect portgrp 45 Name ID WWPN HostType Profile portgrp volgr pID ESSIOport p770_02_lparl_fcO 005C 10000000C9808618 pSeries IBM pSeries AIX 45 V45 all p770_02_lparl_fcl 005D 10000000C9808619 pSeries IBM pSeries AIX 45 V45 all dscli gt Isfbvol volgrp v45 Name ID accstate datastate configstate deviceMIM datatype extpool cap 2 30B cap 10 9B cap blocks p770_vol 6200 Online Normal Normal 2107 900 FB 512 P8 50 0 104857600 p770_vol 6201 Online Normal Normal 2107 900 FB 512 P8 50 0 104857600 p770_vol 6202 Online Normal Normal 2107 900 FB 512 P8 50 0 104857600 boot_LPAR1_9002 9002 Online Normal Normal 2107 900 FB 512 P8 50 0 104857600 dscli gt Ishostconnect login WWNN WWPN ESSIOport LoginType Name ID 20000000C9808618 10000000C9808618 10005 SCSI p770_02_lpar1_fc0 005C 20000000C9808619 10000000C9808619 10005 SCSI p770_02_lpar1_fcl 005D Chapter 5 AIX considerations 63 5 1 2 Using node port ID virtualization NPIV When using virtual Fibre Channel adapters two virtual WWPNSs are assigned to a given virtual Fibre Channel adapter by the IBM Power System Figure 5 1 shows the properties for a virtual Fibre Channel adapter on a Power HMC GUI r Virtual Fibre Channel Adapter Properties ITSO LPAR 5 M 9 155 50 104 htt
427. x To remove a disk device or volume from Linux and avoid system suspending due to incomplete I O requests follow this sequence 1 Stop all applications that use the device and ensure all updates or writes are completed 2 Un mount the file systems that use the device 3 If the device is part of an LVM configuration remove it from all logical volumes and volume groups 4 Remove all paths to the device from the system as shown in Example 6 38 Example 6 38 Removing all paths to a disk device xX36501ab9 echo 1 gt sys class scsi_disk 3 0 0 3 device delete x36501ab9 echo 1 gt sys class scsi_disk 3 0 1 3 device delete xX36501ab9 echo 1 gt sys class scsi_disk 4 0 0 3 device delete xX36501ab9 echo 1 gt sys class scsi_disk 4 0 1 3 device delete The device paths or disk devices are represented by their Linux SCSI address see 6 2 5 Linux SCSI addressing on page 94 Run the multipathd k show topology command after removing each path to monitor the progress DM Multipath and udev recognize the removal automatically and delete all corresponding disk and multipath device nodes Make sure you remove all paths that exist to the device After all paths are removed you can detach the device on the storage system level The watch command Run the watch command periodically to monitor the multipath topology To monitor the multipath topology with a period of one second see the following command wa
428. xceed that number or where the queue depth on certain devices might be increased to more than the default additional adapter slots for the VIOS and the virtual I O client partition need to be created The VSCSI adapters have a fixed queue depth There are 512 command elements of which two are used by the adapter three are reserved for each VSCSI LUN for error recovery and the rest are used for I O requests Thus with the default queue depth of 3 for VSCSI LUNs that allows for up to 85 LUNs to use an adapter 512 2 3 3 85 LUNs If a higher queue depths for the virtual SCSI disk devices in the client partition is set then the number of LUNs per adapter is reduced For example with a default queue depth of 20 for a DS8000 LUN as backing device using SDDPCM see Example 4 10 you might need to increase the queue depth of the virtual SCSI disk device in the client partition also to 20 for best performance For an AIX client partition it allows the following DS8000 LUNs per adapter 512 2 20 3 22 The default attributes of a DS8000 device on the VIOS with SDDPCM version 2 6 0 0 are shown in Example 4 10 Example 4 10 Attributes of DS8000 volumes in the VIOS partition using SDDPCM Isdev dev hdiskO attr attribute PCM PR_key_value algorithm clr_q dist_err_pcnt dist_tw_width hcheck_interval hcheck_mode location Jun_id lun_reset_spt max_transfer node_name pvid q_err q_type qfull_dly queue_depth reserve_
429. xpert Partitioner panel Figure 6 5 enable multipathing a Select Harddisks in the navigation section on the left side b Click the Configure button in the bottom right corner of the main panel c Select Configure Multipath Expert Partitioner System View Hard Disks Device Size F Enc ype ai FS Type Lat Bidevsda 136 62 GB ServeRA Dih idevisdal 39 99 GB Lin idavisda2 36 9 GB Pan i Bal Volume Management jqayisdaa Abort Configure Multipath Configure u Bj Hard Disks ide if Crypt Files tell Pewee 4 gt Rescan Disks Configure Figure 6 5 Enabling multipath in partitioner 120 IBM System Storage DS8000 Host Attachment and Interoperability 4 Click Yes when the tool asks for confirmation The tool rescans the disk devices and provides an updated list of hard disks that also shows the multipath devices it has found as you can see in Figure 6 6 Expert Partitioner System View 2 Hard Disks S T a n N a Be em linus Device size F e bE Hard Disks idevimapper 3600507630atfc29f0000000000004690 50 00 GB oy RAID dev mapper 3600507630affc29f0000000000004691 50 00 GB Fe dev mapper 3600507630affc29f0000000000004691 partl 50 00 GB j Volume Management sdev mapper 3600507630affc29f0000000000004790 50 00 GB E crypt Files dev mapper 3600507630affc29f0000000000004791 100 00 GB it ae dev mapper SServeRA_Drive_l1_2D
430. y dev disk disk14 Physical volumes PV Name dev disk disk14 VG Name dev vg01 PV Status available Allocatable yes VGDA 2 Cur LV 1 PE Size Mbytes 4 Total PE 4607 Free PE 1607 Allocated PE 3000 Stale PE 0 IO Timeout Seconds default Autoswitch On 196 IBM System Storage DS8000 Host Attachment and Interoperability 5 Extend the size of the LVM logical volume by using the 1vextend command as shown in Example 10 16 Example 10 16 Extending the size of the LVM logical volume Ivextend 1 4600 dev vg00 1vol1 Logical volume dev vg00 1lvol1 has been successfully extended Volume Group configuration for dev vg01 has been saved in etc 1lvmconf vg01 conf 6 Increase the size of the file system and display the mounted file systems as shown in Example 10 17 Example 10 17 Increasing the file system size fsadm F vxfs b 18400000 vmnt vxfs fsadm V 3 23585 dev vg01 rlvoll is currently 12288000 sectors size will be increased bdf Filesystem kbytes used avail used Mounted on dev vg00 1vol13 1048576 312368 730480 30 dev vg00 1vol1 1835008 151848 1670048 8 stand dev vg00 1vol8 8912896 3442712 5430112 39 var dev vg00 1vol7 3964928 2844776 1111440 72 usr dev vg00 1vo14 524288 24136 496304 5 tmp dev vg00 1vol6 7798784 3582480 4183392 46 opt dev vg00 1vol15 131072 123136 7928 94 home DevFS 3 3 0 100 dev deviceFileSystem dev vg01 1voll 18420000 21603 17248505 0 vmnt Tip To
431. z196 and z114 is supported by DS8700 and DS8800 starting at LMC level 5 1 This function was developed to reduce the complexity and skills needed in a complex FICON production environment for changing the I O configuration With zDAC you can add storage subsystems to an existing I O configuration in less time depending on the policy that you defined ZDAC proposes new configurations that incorporate the current contents of your I O Definition File ODF with additions for new and changed subsystems and their devices based on the policy that you defined in the Hardware Configuration Definition HCD The following requirements must be met for using zZDAC gt Your System z must be a z Enterprise z196 or z114 running z OS V1 R12 with PTFs or higher gt LPAR must be authorized to make dynamic I O Configuration ZDCM changes on each processor hosting a discovery system Hardware Configuration Definition HCD and Hardware Configuration Management HCM users need to have authority for making dynamic I O configuration changes As the name of this feature implies ZDAC provides two capabilities gt Discovery Provides capability to discover attached disk connected to FICON fabrics Detects new and older storage subsystems Detects new control units on existing storage subsystems Proposes control units and device numbering Proposes paths for all discovery systems to newly discovered control units including the Sy

PDF - IBM Redbooks

Contents

Download Pdf Manuals

Related Search

Related Contents