Chapter 1
Contents
1. Domain hame Sarr Status Disasled Print DHS bere Current primary clustered system is standalone and secondary clustered system i shutdown Date replication j not active duc to clustering i mot in narmal mede y Por Helo press Ep Local Cluster Type Primary rules Status F gt The first screen presents the default setting of the cluster server which sets the Failover Failback functions disabled The cluster server is now in replication only mode When in replication only mode the Primary and Secondary stations are fixed with their status that is once set Primary a server will always hold the Primary status in the cluster system The selection for Primary and Secondary stations can only be done when Failover is disabled This setting is used for replication only mode only to determine the data flow direction Step Iz Seitchower Events Setup 1 Automatic Failover Event Setup Enable laver Ense dats labok mihol or Eoo zem Does greeted Fe Remit ranpa dudes Mees dive vg pei e Fed io stort adower Inggenng eyed Pub mang che etek adepter fers fo function sable adcser cni 6 Local scien is the prima stator C Local spslem is the secondasp station 31 By select Enable Failover ClusterReplica SQL 32. nol be an eating hiti nase on po cate ele 11 reca ak Boe e ue Fase the cluster nama anseres lha ere ira og Go oe nama lo ihe eonfigurason tit rcs ca Mis Eee Loos ae ied Sree cn pour payant 4 Enable michon chele rame Churcher He of enaractisis taniy Mane gt The next step is to define the cluster IP This is the IP that is to be attached to the Primary station of the cluster server for network clients to connect to the server If you have more than one network cards in the system use the drop down menu to select the card with the IP you wanted If DNS setup is configured this cluster IP can be optional Step 3 Custer Address Failover Setup n X Cluster Address Failover Setup Mies ee a a Rider airi viuia addosooss well moved bo fre nev pma ipiam bo conta ha ingas Ip the chariy v Enable IP address bstorer Select banding meted cand In accion aiddiess For public access Reate BILE 33 27 de Fonds Fact Ethernet HIC Property IP addicts l on addesser ans virtual foi herds ln Serat be hin need which releck cand ihe visual adderat vel be binded rou can click on Ping button bo test F the speciied adders hos been usad 195 2 1 7h otra 258 258 D LT Remove
3. 8 14 22 28 295 14 49 29 95 8 1450355643 1450135543 14 50 44 052 1450 42 054 8 14 50 55 718 amp 165028 719 amp 14551 09591 Daa Dats Data Aene Dials Dials Dats Daa Regn Daa D als D sts Herc Data Respe Data Dials espre Daa Respe Data Respe Dala Dats espre Das Resme Sial egre Tor CS Program process Ciro m Fre Sial ies process for L Progesm Fi Aese process C Program Files Sial reduc process Thor CProgeam Fi Hasmie process bor rogram Files Start resumo process for C Program Fi Resme process Program Files Sial eare hor CAFen Fi Rasme process bor Crogan Filezs S arl process for C Program Fi Arane process iu C Program Files eare process For CS Progeam Fi process Ciro em Files process or L Program Fi Poesynic process bor C Program Files Slant resur hor Program Fi process for ragam Files Stan resume process for LS Program Fi process C VProgram Filezs 14 3 2 Other Cluster Status During Operation Following lists some important cluster status when the cluster server is in operation with simple explanations 1 Cluster server is in Always confirm mode This mode affects only the fil
4. X Beira CURA ete Fea Hips orm Loca Cipkber err E Ful Gere Sate dier 4 Following picture shows the cluster server is in Failback process During this process files on the standalone server are copied into the newly start server and upon completion of this file sync the newly start server will be reset as the Primary station a Chaska Sacer paar Stat E EO osi id chri 21 1 Chester Status qe aas Tren CTE 12 838 Chester shal s s rw ains uer o i maria siris Geek lha srm en b gri chet ones es Hp Meda ates ahd coda 4A Ceid P Acher 198 2111 BD us Ten nns 1 PM m I P das ded Ai nee de Red Mn ein bebe ami fe eed curiaung reram Fie brieg replicated CAPregram 5040 Sena sot gpp Lap LDE ROCA oe UOIne2r MWy XA Pasian Ort aee Fea Hels mmr FL Lord Cipker Tams Pl Be Sebo poy mam 16 5 Following picture shows that the cluster server is waiting for the systems to swap roles This status of the cluster server can be found in one of the situations e User manually switch the role of the Primary and Secondary stations by click the button Switch Role on the status screen e At the end of a Failback process when
5. e Compare Replication File Info MEE Ee El al xj Biter Exit C Program FlessMierasolt SOL Servers S SOLND ala Icd 4 Canespendng Replicalian Path on Secondary System CXPragram SUL Serre S S OLD ara Last Modified C Program Files Microsoft SQL SereersddSSi3LNDatasdisimd Identical 736432 pe Dem Do 87 13 C Program SOL Iderbcad 2359296 pe Dem Do 87 13 Program Fides Micraso SOL Serwer MSS OL YD ayva lderhcal 118993088 W265 11 34 48 C Program Microsoft SQL Seren MSSL atavmasio Iderhcal 96430 4 2565 11 34 48 MPragram Files Microsoft SQL SereerMS SQL ND ara vmodel Idenhica 65535 DFES 11 34 42 Program Fides Microso amp SQL Seven MSSL Identical 524238 11 34 42 C Program Fides Microsoft SQL Sereen MS SOLD ara vmsdb Identical 11532335 D44 25 065 11 34 42 C Program Files Microsoft SQL SereerMS SOLD Identical 2359096 11 33 44 C Program FdezsMicrosoft SQL Serve MSS OLD alaner ldernhca 10240 4 06 05 16 5E 15 C Program Fides Microsoft SQL Servers S SOL NDS Hen Iderica 10752 DATE T5 16 5E 13 C Program Fdez Microsoft SQL SereerMS SOLD north Idenica 1096576 D4 2 05 11 33 41 C Program FidesMicrosof SQL Seres S SOLDI Snoth Identical 2752512 2 05 11 34 41 C Program Fdes Microsoft SQL Servan S SLM aS spubs Iderhes 18443555 D4 75 15 11 34 41 C Program Files Microsoft SOL Served S SOL SD atespubs Iderhca 7
6. 4 22 Advantage and disadvantages in each of the Failback method Advantages e The automatic Failback is simple and hassle free to always keeping the original Primary station to be the active server Some users prefer this way because they use a heavy duty machine to run the Primary station but only use an average machine for the backup Naturally they would prefer to use the original Primary to be the active server as much as possible Automatic Failback will without confirmation first copy all current data on the standalone server and then switch the roles e The confirmed Failback guarantees no mistakes on data overwritten In some special circumstances the newly started server may contain the data users wanted Users then may want to copy the data on the newly started server to the standalone server instead of the other way around as normal Failover will do Manual Failback gives you the full control of in which direction you want the data to flow 24 Disadvantages e The automatic Failback In some special situations the newly start server may contain the data users want to keep If automatic failover is set on the cluster server this data will be overwritten as soon as the system is connected into the clustering This accidental loss of data then becomes inevitable then e The confirmed Failback requires a person to sit in front of the server and make decisions on each move of the configuration for the right result
7. Pes Back Nest gt Cancel 33 Ifno DNS setup 15 needed leave the box unchecked If DNS configuration is needed following conditions require special attention Step 4 ONS Redirection Setup Hep 4 Prefered Replication Role and Replication Methad Eg ai nad data charge in vll br perl bo pho ib peed eh caer i seleched kcal dala vall be ihe repica ol mater lation IClusterF episc a eller bn seplicabon mode repkcalien apples daba hargas dente Lo Lee menor dinachion on Sabor AF voribored Vall be giard po stopped on seplica tation bo allis changes access repbicalion sons all dala changes m a langaia bon sepbca slation ard ork upises Langehad dala accondg bo Linge miena or intel ever Hepi accass mode ales bed bo query and gee abe riore yay pl lh te ANOLE CIR domam controller or DHS Pins DNS Siero 123 123 123 1 Secondary DHI 155 l 73 Ful DNS name CuceHepeca sc osi con Local IP fei 7 i Fiemote IP for redirection i The most important condition being that the clustered systems must not be the domain controller or the DNS server Other issues involve DNS configurations are 1 Users must be able to authorize DNS dynamic upda
8. This process can be tedious and cause error by human mistakes 25 CHAPTER 5 Cluster Configuration This chapter provides details on concept of clustering in ClusterReplica SQL 3 0 and the configuration procedures The two sections in Cluster Configuration are 1 Replication Setup deals with the data replication communications between the clustered servers 2 Failover Setup deals with the public connection of network clients to the server and the cluster failover between the clustered servers ClusterReplica SQL 3 0 binds two Windows systems together to make up the cluster server During normal operation the Primary station is responsible for servicing the network clients and replicate the new data to the Secondary station in real time so that all data on the Secondary station will be current within minutes The Secondary station while remain hidden from the network clients 1s standby and ready to take over the active role should the Primary station fails for any reason Meanwhile managers can extract data from it for routine checkup or making periodic reports without interfere the public service job of the cluster server 5 1 Replication Setup This part of the Cluster Configuration configures the cluster server sets data replication and file sync routines A little different from Failover Setup Replication Setup allows some items of the configuration to be done on both Primary and Secondary stations To be specific the
9. backup The default temporary workspace location can be changed by click Advanced button emporarg Workspace Path 44 CHAPTER 8 Other Utility Tools Along with the ClusterReplica SQL 3 0 comes with some utility tools that may help users better work in their network environment 8 1 Network Utilities There are times when a user needs to know the IP address of a registered host or the network transaction routing path ClusterReplica s Network Utilities tool can help you achieving this objective This network utility tool consists of two features Host Lookup and Trace Rout gt Host Lookup Host Lookup can help users match the IP address of a web host with its name As shown in figure 6 1 below to find the IP address of yahoo type yahoo com in the host name window and click on button Look Up The IP address for yahoo com will be found and filled in the P Address window i Network Utilities EC C X Hast Lookup Trace Route Please enter IP address c Host Name to lookup Host Hame yahoo com IP Address 216 115 108 7 A reverse lookup can also be done using Host Lookup In this case user simply types in the IP address to find the registered host s name Trace Route 45 gt The second feature in Network Utilities is Trace Route This tool is created to help users trace the routing path of service providers between the host and the traced servers Up to
10. is s Pe E PE Dur B 1 Set this machine to be Primary ab M M 7 Inthe next screen some IP addresses are to be filled in NOTE The IP used here is for the cluster server s internal communication Define Private Network Information ro Setup private networking informalion ue T E The private nehwork establishes rode to node commurscalion Ea asg cluster status signals and cluster dat 3 Network adapter of private network Sires car Selected adapter will be used for intemal clustered 1 systems nage _ ROIS BO iver _ DE 2 i 25 Private contol IP addiess E ITEA Enter new private control IP address at browse from existing IP _ addresses on the private network ad qu TE 2 d Local private control IF adess 123 123 EE 16 z Browse E Loca 4 pivale mask _ 255 F 268 2068 0 js cies cu UNE Spec he conl of e her hte node vr Loo For one network card systems the internal and external communications are using the same subnet IPs there is no need to distinguish For systems with two network cards the internal and external communications are using different subnet IPs be sure to enter the IPs that are to be used for internal communications As in above example the 123 subnet is entered We reco
11. 20 router activities will be listed a Network Utilities T E 6 HostLookup Trace Please enter the IP addes you wart lo race x x Resolve IP to hast name Tireoul for each reply ho me Maximum number of router court Trace seus Host Address It is fairly simple to use this tool Select Trace Route tap from the Network Utilities GUI type in the Host s IP address and then click on button Trace The packet routing path of all internet service providers up to 20 will be listed in the Trace status screen 8 2 Network Application Scanner The Network Application Scanner can scan the operating system and find out all network applications running on it From its dialog box click on button scan to get the services of all applications currently running on the system listed M sd fut De E IE 18 18 18 IE 18 18 E E G 4 mms genas ee exe Diepi Depienne mati pos exa 5455 EXE SERYVIDES ERE Sustem System Suslem Sustem System Suslem System Susiem Sustem 46 Click on button Details will get a popup box to come up showing you the detailed information of the network application Process Info x Proces Inforneaticer Process Hame SERVICES Process Path epstem32 services exe Creshon Tene 12 7 1999 File S
12. C Program Fieri amp 051018430 Oats Reyne Stat rait process for C Program Fi 051113265 Fennec Repne process hor C Program Fier 06 11 19 255 Daa Stat rayne process for C Progeam Fi amp 0511256558 Date Reme process For LXPsogram Fides T2659 Daa Repe Stat pectine pocast for C MProgeam Fi 051128551 Data Rene Reang process for C APspgeam Files amp 051128561 Daa Renne Slat rayne process lor C VProgram Fi 6051149131 Reame process for C Program Fiesi amp o51810159 Date Hemi Stat rest poces For CL MProgeam FL i b If a Event Description 1s too long double click it will bring up a message box with the full description listed in it as shown in the picture below Eve nt Desc ript ig Time 06 03 09 144 Type Currant cluster state does not allay to recene command Feplicalon request is aborted 41 Scheduled File Backup This tool completes the data protection routine by assisting users to backup the data on the Secondary station to a remote location Why the Secondary station because the backup process will not affect the server performance on the Primary station which services the network clients during normal operation When select Open Data Backup Utility on the main menu the Backup configure tion panel comes up on the right As regular backup configuration you need to provide e Backup source e Ba
13. IP idees 128 21 11 D ati anii ie wei iie i rail Low QU B een ev eal del a 2 me e t ES e Caneal primary duster sytem ard accom dary closlerced is welling fer s ceirean Date replications Hi aer active due Ya io ial in mode Bg Fes Help orm local Dasher Frane Jd Pul uero Shue 17 7 Following picture shows a Holding status on the Secondary station This status indicates that the Primary station failed during a data replication process and that the data on the Secondary station 15 incomplete The Secondary station is now running in standalone mode but it cannot take over the active role because it does not have the usable data to server the clients Two things can be done to correct the situation 1 Manually restart the Primary station to perform a Failback without having the data on the Primary station to be overwritten that 1s make sure the Primary station is in Always confirm mode see status description 1 2 Fix up the data on the Secondary station and switch its role to be the Primary station This condition is set for data protection purpose 79 opc ard d Lend Irira J Cesar 1525 2 1 1I Chuster Status e E E 13354 Chium aha gs Sew nmt iat rra er cime Chee eed
14. completely out of the cluster server and the Secondary station is to operate as standalone server until the Primary station 15 plugged back to the cluster server Following pictures present the situation graphically where the two systems in the yellow frame make up the cluster server 1 Picture 1 shows a normal working situation of the cluster server 2 Picture 2 shows the situation after a Failover Case 2 Monitored service malfunction or fail to start In this situation one of the monitored services is detected of malfunction but everything else are working just fine ClusterReplica SQL 3 0 on detecting the service error will automatically switch the roles of active standby stations while both systems are still running Following picture give the general sense of the situation 3 An Example The SQL service on the Primary station stopped for some reason However all other applications are working normally 21 ClusterReplica SQL 3 0 detected the failure of SQL service on the Primary station It starts the SQL service on the Secondary station immediately and checks on all services in the monitored service list making sure they are also started on the Secondary station At this point the Secondary station takes up the active role and the original Primary and Secondary stations in the cluster server switched positions This switch of roles 1s done without network clients notice The system Admin can then work on the st
15. new data ClusterReplica SQL 3 0 will force a break on the data in flow and perform the data transport at every ten minutes 3 Temporary workspace path this location is where ClusterReplica store temporary files during operation If large database files are to be replicated and report mode is activated a large space for this temporary workspace may be needed Users can change the location to a different drive for this consideration 6 3 Replication Verification Schedule ClusterReplica SQL 3 0 can run automatic data replication verifications which will auto update files on the Secondary system whose file dates or file size does not match that on the Primary system The default set on this data replication verification is as shown You can change the verification schedule according to your needs Verification schedule setup E E XI Dats venhcabon schedule Venicabon scheduling venies and mantar data accurancy and masimizes replication performance E very dais 1 00 00 39 6 4 Compare replication file info The Compare Replication File Info tool allows you to manually check up the folders and files in the File Replication definitions to make sure all files are copied properly from the Primary server to the Secondary server When in File Replication Setup click button Advanced on the menu bar to find Compare File Info to open up the processing screen as shown below
16. only item that cannot be changed on the Secondary station in this part of configuration is to set SQL operation mode This design requires special attention to following 1 The communication port and communication addresses must match on both Primary and Secondary stations Changes made on one system must also be done one the other system 2 The keep alive time is the length of time by which a system determines if the other system is still alive Because this value is system and network traffic dependent it can be set of different values on the Primary and secondary stations 26 3 The Data overwritten protection can also be set individually on Primary and Secondary stations When the check box is checked it means at the restart of the local system confirmation on overwritten of local data files will come up before action is taken v Wink CherbesR eplca SOL Edtion 3 8 Hode for Fayen Enabled Setup SOL Operaben Mods On Secondary Sete el Diable 5UL sereer operation on zeconcan sehen Poe deect Schedule Keep Ades Times Keep alive Gina 15 seconds r Eoniananic lion Pot ameb PIM comnsunscalcn rides pair i sesenved ici conimanicsiion aed replication fom paran 5 Bising sclacia DET Free Current primary clustered system Is standalone and secondary clustered system is shutdown Data replication is not active
17. services fails Failback is the process in which Primary station resumes its Primary role after an Failover event 4 1 Failover or Not Failover The main purposes of clustering servers are 1 Real time data replication for data protection 2 Failover for server 24x7 availability While the first situation requires no Failover function and the second situation does ClusterReplica SQL 3 0 supports both purposes 4 11 Using the Failover functions To use the Failover functions of ClusterReplica SQL 3 0 users need to configure Failover Setup with user defined cluster name and pre set cluster IP see next section Failover triggering events 1 Receives no response during keep alive period The Secondary station monitors the active state of the Primary station in a pre set duration of time The default setting of the keep alive period 1s 15 seconds Users can make adjustments to this time period for ClusterReplica SQL 3 0 to determine when to Exercise Failover Monitored services malfunction or fail to start when any one of the monitored services defined 1s detected 1n malfunction failover will be activated Public connection network adapter fails to function when the public connection to the Primary station fails either due to network card malfunction or the disconnection of the network cable failover will be activated 20 Case 1 Receives no response during keep alive period In this situation the Primary station is
18. siris ue oo lha scr Ee gri POLL cael GL nm asa D d Floral pps a Coated IP iae 121 11 otim when rc ie oes kh BE a Chei miii dr aui Fa raced curtain raram E Larreal piner duki system ics aad cuslered ia anren Data replication bi aer actiye due Ya clustering is malin rmt male amp Pot Hep orem ll Th teed Cite Pere E A z gt 8 Following picture shows the Failed Server status which indicates that the ClusterReplica SQL 3 0 service on local system is not running properly It now has lost connection to the remote system and therefore cannot decide the role it must take in the cluster server Users can try to restart the service or reboot the system to bring the system back to normal state 18 T te mag Wm Hip Ee pp Cluster Status X Crean P Pace ana icm 7 x mr Tre Chine as ales Leur mo e T khe ned ed rias Lisa umbra n Dkk lha apri Eo gel HEEL dosciptions d Nei x bun Und X Dinkel iF red ram QD numum wem ie em tede m RUM pe cma Sei P da uOIe9e22 n M Carrea
19. the terns ok Agreement IEEE T _ og UU crete E Eu This software peegram ard documentation aie zwar nol sod to you You havea ue ail the teams of tha ecu cA the a p 2 A wit Sess ducite dene SQL EERO you epus ied di x 05 iuda i Please read the agreement carefully and be sure you fully understand every term listed and that you agree with this agreement entirely before proceed with the installation Click No if you disagree with any of the terms listed in the license agreement The installation process will abort If you clicked Yes in 3 the next screen prompts for you to fill in the license key for using the product Installshield Wizard Ea License Key Dialog The Sr pass needs lcense key ta continue icense Te ue bow iy oe pa cick NEST AGES MM m MM The License box 15 filled with DEMO as the default entry If you have not purchase a license for the product and just want to install the demo package leave this default entry in the box and click on next to continue The next screen asks you to decide if you want to set the local system as the Primary system or the Secondary system Select one and continue Installshield Wizard Ea Setup Type Lhooge the sehup type that best suits pour needs ar i Set the current machine as either the primary server OF the Eee server
20. to not failback they can check No in the confirmation process With this action ClusterReplica SQL 3 0 will keep the current active station to be the Primary station until the cluster server failover again 22 P 15 an example of confirmed Failback The newly start system is plugged into the cluster server and the dialog box for user confirmation comes up e In following picture Yes switch local system to Primary station is selected If select the other setting the newly start system will remain to be the Secondary station when the file sync process completes Clustering role confirmation E i X Confirm Cluster Role Remote system is currently running as primary clustering Saon you went to swatch local system to be the prima clustenng By selecting this option you are granting local system the primary clustenng role to communicata wath clients and instrucbnig the remote system to be the backup system m standby mode C Ma make local system secondary clusbenng station By selecting Ihe option you ae making lacal sysbem the sezondan elustenng elation which wil act backup spiten in stande mode e Inthe next picture you have the choice of having the data on the current active standalone station copied over to the newly start server or not having it copied over In most common situation you would want the data to be copied over because the active server should always ha
21. xi IT Coated Plc 28 2 n K ree an D n EE Ls nei HN aan e Sebold D gia hp Bar BER Lj Chae Hina cheated Dir Akbar Sor Chive red ote aed dary 0 cystem is slandbey replication 1s currentty X CURT Wc Pos Help orm Fl fh Lond Cintas rrer Ful Caere Soins dia 15 3 Following picture shows the cluster server is to prepare for Failvoer It means that a malfunction of the Primary station is detected and the Secondary station is getting ready to take over the active role and running in standalone mode EE P d Local Scion 4 Cosi E Erbe 125 2 1 181 Chester Status Tus CATR Chister sbad s Mee al neis Uer Do e rana er net alate oo lha nsnm scr be gri piai orar m up Meda ate cares n d Annais Soins i Cei LP Aie 1821 11 EL Mu wd i i res Tine id Stau IN e U goa c as Tae Fa vifo ccn ORT RO 7100 EO rJ Chanter eed ned bebe aai alara Le 732112 Low Dur s uu arreat primery dusted active ard arem dary clustered system is F Data reglizabos ks aer Please vite chsterg bg end seplicenasn big foe more Cee
22. 0 enables Failover Failback functions The cluster server 15 no longer in replication only mode Step 1 Switchover Events Setup step 1 Automatic Failover Event Setup f Enable failover Enable automatic data fallback without confirmation for local system local is preferred primary system Automatic failover triggering events Receive no response during Keep Alive timeout period v Monitored service s Fail to start Disable failover replication only Locallsystem ie the primary station Locallsystem is the secondary station lt Back Cancel Set Automatic Failback If the box Enable automatic data failback in above picture 15 checked automatic Failback will take place A warning as shown in next picture should come up to remind you no further confirmation will be checked before data overwritten Leave this box unchecked means manual Failback is in place You currenth have canfirmatian enabled Far date Da you want bo disable confirmation and set auto Faiback Far local system 32 gt The next step 15 to give the cluster server a name for the cluster server This 15 normally the name of the SQL server but it cannot be the Windows system name If DNS setup is configured this cluster name can be optional Step Chuster Same Failower Setup Hep z Cluster Name Failover Setup ble this opting vali add ackflional cheer name bo gos primary autem fet nae
23. 195 2 1 76 1s added to system Steps in details With a successful download you will get the ClusterReplica sql exe file as the installer Double click on it to start installation process 1 The first screen as seen below greets you with the product name et xLink ClusterReplica MS SQL Edition for Windows 7000 XP 20035 InstallShield Welcome to the InstallShield Wizard for XLink ClusterReplica MS SQL Edition for Windows 2000X PF12003 The InstaliShield Th wall help inshall Link ClusberRepliza MS SQL Edition Far Windows 2000 XP f 2003 on your computer Ta continue cick Next 2 Click on button Next to move on Following message box will come up to remind you the software requires a member of administrators group to perform the installation Thie prostant tesis amber Merino Group ta ss haves to continue octo qut the setip Mo riy mide Ehe pu REG ante canta USE EUR Ee Te 3 Ifyou are already logged in as a member of the administrators click on yes to continue Along the way you will see the license agreement with conditions on using the software properly Installshield Wizard Ea License Agreement Pisae read the foloveng license agreement carelulu Press the DOWN kep to see the rast of the agieemer 00000000000 SEN fou are authorized to access ded i you us OR on ent vih Hink Ine i use of the to
24. 2 4 Removing ClusterReplica ClusterReplica MSSQL Edition is equipped with its own uninstall program You can find it by clicking on Windows Start Programs Xlink ClusterReplica If for any reason you want to remove the software from your system run this uninstall program will remove all files of it Using Windows add remove program should reach the same result Be sure to reboot the system after running uninstall 11 CHAPTER 3 Understand ClusterReplica SQL 3 0 Status The successful installation of ClusterReplica SQL 3 0 sets the cluster server in replication only mode The default settings of the cluster server is presented below 3 1 Checking the Status on Your Cluster Server In this first screen we can pick up following information e The server is the Primary Station e The local system has Control IP 123 123 123 10 and functions as the Active server e The remote system has Control IP 123 123 123 16 and functions as the Standby server e The scheduled Data Verification is enabled e The Failover function is disabled Users need to configure the Failover Setup for automatic Failover and Failback functions Hm di m d Amieen Fake Pury orh n hot Configuration Ru n rpm usc Cluster Status J 2L e Teak Fare DEL 1157110 Chester stalus ew aloe Wear in MERIT cuim serie x ach ick an
25. 36432 042263706 11 34 42 C Program Files Microsoft SQL Iderhca 8368506 D4 26 065 11 32 07 C Program SOL Server MSSOL Oatevempl Identical 524238 D4 26 065 11 37 07 Far Help press F1 Di ferent 0 4 The first thing to setup here is how you want the comparison result to display Click the button Filter on the menu bar to get next dialog box Select the display option according to your needs and click OK The default setting is on Display all results File Comparison Option T 9 C Display only the different ones C Display only the identical ones 40 The next thing to do 1s to find the source file of the replication Browse by click Click button amp get the comparison results By default all files and folders are to be listed If there are problems in some replicated files you may find wordings such as file does not exist in the Remark column 6 5 Data Replication Log The data replication log is created for you to trace the replication actions on files or folders and decide if some files or folders are not properly replicated that you need to run a manual sync to ensure accuracy of the data on the Standby system Data Replication Log mm 1 QE m 1 144 D He T C LATET al un doc nob d o tor amp o5 s5553 Start rsync process for C Program Fi 05 10 18 430 Daa Penne Fearne process
26. Table of Contents ClusterReplica SQL 3 0 User Manual bt Gl eee nner te ee nee eee 1 A 1 Led WMhatis ClusterReplica SOL 2 0 1 1 2 Whom is ClusterReplica SQL 3 0 Designed 2 1 3 How Does ClusterReplica SQL 3 0 Work essere 2 How ToGe He a iom teta biete 4 CHAPTER Creta RH MITTIS PII CORDE 6 Installing Removing ClusterReplica SQL 3 0 6 2 1 System Requirements for Running ClusterReplica SQL 3 0 6 22 Installation GIG e a eee toties eina ito 7 2 9 Running the evaluation 11 2 4 Removing CIMSter Re pli Oe ema ia is 11 CHAPTER A 12 Understand ClusterReplica SQL 3 0 Status 12 3 1 Checking the Status on Your Cluster Server eese 12 3 2 Other Cluster Status During 15 Se Eu M 19 CHAT TER ALoaosoxeezoxecutuiuAk bebenePit v a Eder di VE 20 Understand Failover and Failback 20 Zr padoyet OF NOE AONE oeste Mad atten bns 20 4 11 Using the Failover functions ccccccccccccecccceeeeeseeseeseeeeeceeeeeeeeeeeaeaaas 20 4 12 No
27. a Default version when data transactions have been idled for more than the inberval Database verson point lineout seconds Regar version point timeout seconds Forced version poirt timeout is used ta generate safe dala version when data transactions have been busy for more then the specibed interval Forced version point timeout 10 tg T empara workspace path Workspace for temporary storage ENProgram Files NClusterFreplica SL Explanations of the values and their consequences 1 Virtual memory usage the virtual memory allocated here is for ClusterReplica data replication internal process use More memory allocated here the better performance for ClusterReplica data replication However too much memory used by this one application may hinder performance of other applications So balanced setting is recommended 38 2 Replication Version point values included 1 this section are related to how often a new batch of data is to be transported from the Primary station to the Secondary station e Database version point time out this value indicates that data transport for database files takes place at every eight seconds e Regular version point timeout this value indicates that data transport for non database files takes place at every four seconds e Forced version point timeout This value indicates that if there is a non stop in flow of
28. andby station to figure out what went wrong with the SOL server on it 4 12 Not using the Failover functions When use ClusterReplica SQL 3 0 for the real time data replication purpose only the Failover Failback functions are not configured This is the default setting of ClusterReplica SQL 3 0 with the successful installation of the software on both Primary and Secondary stations Users need only to configure the data replication and the backup utility 1f scheduled routine backup is also desired 4 2 Failback or Not Failback The main reason that some users want to Failback 15 that of the two clustering systems the one used for the Primary station 1s more robust then the one used for the Secondary station Failback will keep the more robust system be the active server as much as it can Notice that if the two systems the cluster server are comparable in hardware components and configuration let the Primary station float between the servers without Failback can be equally efficient 4 21 ClusterReplica SQL 3 0 offers two ways to Failback Automatic Failback Failback is to take place as soon as the newly started server 1s reconnected to the cluster server No human intervention 1s required with this Failback method 2 Confirmed Failback The newly started server will first take up the standby role to allow current data to be copied over then failback Users will be prompted to confirm each step of motion If users decide
29. ces to be monitored for failover Advanced Replication Configuration m5 l EN 1ni Fiepkcsion director w Dreckery on Prap 5 Jt Program fier Momoh SOL SL BACKUP Program FlleiiMerioft SQL Senes HS SOLVEACKLUP Deme Fig SQL Server S SOLVDATA CUr am Fins M erosoft SOL Server HS SOL Seree iM S SOLVE TDATA SOL Serve HS S GLUFTDATA Fier Micii SOL SOL D Crogan Files Microtel SOL Sewer S SOL VL DOG Filed Micon SOL Servers CAPTO ae fenar SOL Servei s S CL Sy HCL FE acs OF To Delete Service Slaton The uig Tor Dice glo Hore Dates ie HSSOLAEAVER Service In detail these sections can be described as following Data Replication Setup Add all files and folders of files to be replicated from the Primary station to the Secondary station here If a file requires an application to access the application service must be added in the Monitored Service section Registry Replication The Registry Replication functionality meets the needs of some server systems which require certain dynamically changed key definitions of registry matched on the Secondary server for a successful failover Warning Do not replicate the entire registry replicate onl
30. ckup destination e Backup schedule This backup tool also provides the FTP data transport method to promote remote backup to higher efficiency Beton Pook rn vem En Es y comp oo Hackup uhm Leg ENTER x mide n lE E CL bari a inlopen Lint Rebii mh did Dads Package Dei Oar Pa Feld SO Dean bap z Emire eie barip iy FTP pim E Fa Hep pa F1 Local Cur rra Paepe Bato di The four buttons on the top of the configuration panel are 1 Config click this button to setup the backup source destination FTP choice and schedule 2 Backup click this button to perform a one time instant backup this action should not affect the pre set schedule 3 Advanced click this button to change the default temporary backup location if larger space is expected to be needed 4 Log click this button to view the backup actions logged 42 7 1 Two examples on backup configuration e Backup is scheduled at 9 14am on every weekday and FTP is enabled to backup the files in the source location to a remote location The FTP remote directory is the home directory on the remote system that the local system would normally connect to when using FTP protocol a y TI Y SHINE some FTP servers are case se e momen umm mmi mom 43 e Backup is scheduled on every day at 12 12pm without the FTP remote
31. ds excl on cluster spaan pou need speci which ceed binds the sached address Both kcal and remote commuricalian hare fo be valid lor ClucteAlepice lo establish Prou cen cick on button bo best if the specified sddesss is reschabled Select bending rete cand EET gt Loc lcomewrcelon addiessi 123 3123 129 10 123 123 13 16 28 gt The next page contains two sections of configuration Step 3 Data Chverwritben Confirmation and SQL Operation Sebup Step 3 Date Ovenwritten Option and SOL Operation Usai need lo spegi local dala opion at sailing lime appkeable lo Fadaser enabled setup only and SOL operation method seconde suzbeen When bosal rs sel bo eonim dala action al system retailing ime user vell be green the option io handle local dala gt Diets overeiten contimetinn at sustem restarting lime Always conum beloa data k ovenaditien by replication pies s wher ehem pers Laits Applikabbli ta laios enabled made anb Sela an SUL sever operation mode on secondaeg Enabled made Update data avery 2 ie 1 Data overwritten protection this option lets you decide if you want to be prompt at every restart of the local system for file overwritten Check the box means yes pleas
32. due in clustering is not ln normal made il Eapliralisr Slates Por Help press F S Local Chaster Primary Pull Respre Status idle F Click Modify to begin the configuration process Any change involve data replication must go through the entire process to complete gt This first page shows two items Step 1 Clusber Private Communication Port amp Keep Alive Timer Setup Step 1 Cluster Private Communication Port amp Keep Alme Timer Keep abra bma used by ClusteiFieplica ba track the of the ether nodes m the responess recetved vati the keep akve bine lame lacal system val booms standalone node m Bre cluster Clustet pirvele communic lon port m used by ClusteiFienica sallviate bo maintain communication and data on bolh and sscendar sushems li comm mesion port is being used oder applications on either of the Clustered syshems it has to be changed for ClushaReplics to function proper Cluster private communication post hes ip be She same om both pomarm and seconday system ou can chick on Pod Scanner butten ba frd gul Ehe speci wed port is beng usex by onp applicalian aiva timer 15 secondis Donmmunication Fort 2002 Defauk E eara 27 l Keep alive time this 1s the time which helps the local system to determine whether the remote system is still alive The default time durati
33. e sync process when the cluster server Failback to avoid data loss should the standalone server start copying files automatically into the newly start server This configuration can be set in Replication configuration LE Il 4 FP 12682 1 18 Cluster Status St rer EN 13 TIRE Cluster dizd is sine aine ipu to cun p ele Bl hates suaa on lha ssam ccr bo curl description m d Reeds ovine Caid Pacer _ Fete 2 D uat L Vers D ee Lan Ls Chunar EOL m ata e D rao visi acini 2100 8 d Cusen Hon neg A Chire 7098 21 12 s UOHBON oH z a Curreal primary duziemd ILE and zrena amp ary rhaine eyelem le gini ifa h active A Packrgken Pea Help orm Fl fh Lors Cir Tes i Ful Gere hie die 2 Cluster server is set in Overwrite mode This is the reciprocal mode of the first one So when cluster server Failback files on the standalone server will be copied to the newly start server without further confirmation c fae q Gardi Pies 1252117 Cluster Status e fug iren CNTR 218 at lier aine ier Do mr umet ei ria Bf hasty sam PE UE Min a lba agger seco bo curd Rien c Eid pasas descriptions
34. e ask me before any file sync takes place Leaving the box unchecked means no please don t bother 2 Select a SQL operation mode the two selections sets the SQL service on the Secondary station in Disabled mode or Enabled mode In enabled mode the Secondary station would allow users to access and view database gt This next page is the summary of the user configuration If more changes are to be made click Back Otherwise click Finish to complete the configuration process Chester Communication Setup Summary Step 4 Summary Quster local piste communicahon address 123 123 123 10 adaptar HOGS 5 0 divar piale communication addesssc 12412312316 Update data every 2 heuill 29 5 2 Failover Setup This section deals with the setup of the public connection and the cluster Failover settings Because some of the configuration selections depend on users system environment some general consideration of network design are listed below for user reference 5 21 Network Configuration Concerning Failover The three concepts that involve Failover in ClusterReplica SQL 3 0 are Cluster Name the name which network clients use to connect to the Primary station 2 Cluster IP the IP address which network clients use to connect to the Primary station 3 DNS setup the cross network configuration for failover The two types of network configuration established on above concepts that concern clu
35. e should be available 24 7 Following are some examples to further explain the three main features of ClusterReplica SQL 3 0 Example One Automatic Failover In this situation the Primary station is completely out of the cluster server and the Secondary station 1s to operate as standalone server until the Primary station 15 plugged back to the cluster server Following pictures present the situation graphically where the two systems in the yellow frame make up the cluster server e Picture 1 shows a normal working situation of the cluster server e Picture 2 shows the situation after a Failover Example Two Active Active mode When the cluster server is set in active active mode the Failover Failback and real time data replication functions are all working in the same manner as it is in the normal active standby mode Data replicated to the Secondary station will however be first saved into a temporary workspace and copied into replication destination in user defined time interval Here is the summary of the situation e update time interval every 4 hours originally set at 8 00pm e inthe data replication destination on the Secondary station 1 at 8 00am when you make a report it will contain only the data up to 7 00am that day 2 at 4 00pm the data in the replication destination should be updated to include all new data entered up to 3 00pm 1 4 How To Get Help Due to complex system configurations of ClusterReplica
36. file sync is completed and the newly start server is to take up the active role and turn itself to become the Primary station epi SE Pee ME canna seccodu pane EE i ClusterStois Pi Covel HP 195 211 1 e And Tren EFT ERS 12348 mandet cinese Che eed siris Cok Lar nsr er ganan deceptions P Hoda akapa conis E Sots Contre IP 190211 D icti eL gl Chan H aed a feel amp Red min babari de curd hee reached cucina rans Curent primary diuzizmd syrien uakrerem iz ueber ile act adier duc cualerirsg iz mama medr uone2ar W M Fun rate A ICE ee Poo Hates orem Fl Long Serer Full Care Hie 6 Following cluster status will show up when one of the clustered systems is waiting for the other one to determine its role It will occur in a Failover or Failback process Diri s HE Se Feline Aud Piras EE i T Caird Fier 155 2 1111 Chester Status e Sat 1 15 900001 bad ees ipu to moraer cune clustered elis b Casus tik oo Hw nnam Fo descriptions e mue D abest carri n d n Coated
37. he interference of the public network traffic For example system 123 123 123 10 195 2 1 74 system2 123 123 123 16 195 2 1 75 The 123 subnet is to be used for clustered systems internal communication and the 195 subnet is to be used for server clients connection The effect on this setting of two network cards 1s remarkable From security standpoint the separation of the internal and external traffic on the cluster server eliminates the possibility of public interference of the servers internal activities Now data transfer security is ensured On the same footing the improvement of server performance in the eyes of network clients can also be notable First with two subnets network traffic is reduced Secondly each set of IP addresses devoted to specific tasks efficiency increased 2 2 Installation Guide Before install ClusterReplica SQL 3 0 some attention points are listed blow One license key is to be used on both systems for installation A virtual IP must be added on the Primary station on the network card to be used for public connection If each of the servers has two network cards be sure that the two sets of subnets are matching on the Primary and the Secondary stations For example system 123 123 123 10 195 2 1 74 195 2 1 76 virtual IP system2 123 123 123 16 195 2 1 75 The 195 subnet is to be used for server public connection and systeml is to be used for Primary station So a virtual IP of
38. ieg 1 amp 25 330104 1 25 33 656 16 27 00 730 15 27 14 Pod 16 07 35 652 T5 27 90 6602 Wea TE 28 35 6 15 28 43 0026 162903 125 162923145 TE 24 37 165 16 30 11 224 16 30 25 354 16 30 25 483 16 30 53 504 1E 31 13 524 TE 31 33 5613 T5 31 47 Ei TE 32 1 553 localckmbe Fra Locs system is in sarba cue Locs slem iz in se Loss zpsiem is in stante tabe Local zyzlem p in riata ctae Loc re im standby dete Lond m stadbe fale Local p in shale LOCUM piar re wn stancibu stab Local i in standby Loca sedem i im larba hate Loca is ie standby aht Local ssistem k in Aaria state Local gaem iz i Aarbu tabe Local system H in stamdbu at a Loca system is in staria Local sslem is in staba state Loss zgzlem i in standby shabe Local sychem iz in siais state Local m standby state Lass zyzlem rin sale Leal pilem re m stance fale cs Cree rd x I 13 Data Replication Log Action EI Data Re plicatinn Log jj Data Replication Log EH Sep 2005 gl 02 08 2005 File Backup Log backup WordPad Bile Edit View oe Pe Insert Format Help o o Djela ea al gt 58 12 00 53 27 120202 370 122222 170 8120300354 8 120300 354 120308274 8 1202108 272 8120310233 8120310333 8120330507 amp 14 40 08 207
39. ion 46 8 3 Port Reference 47 CHAPTER 1 Introduction ClusterReplica SQL 3 0 is a software solution specializes in handling MS SQL database file replications Running on two Microsoft Windows 2003 2000 XP systems ClusterReplica binds them into one clustering system to serve the network clients 1 1 What is ClusterReplica SQL 3 0 ClusterReplica SQL 3 0 edition is specially designed for Microsoft SQL database servers clustering to perform file replication and failover This special edition 1s created as a turn key product that when the MS SQL server is configured with default settings data replication and failover functions of ClusterReplica are to be automatically configured so that users of not much database server knowledge can use this software without sweat Main features of ClusterReplica SQL 3 0 include Active Standby configuration In this configuration SQL server service on the Secondary station 1s disabled So the Secondary station 1s standby and files on it being updated in real time so that it can be ready should Failover takes place Active Active configuration In this configuration SQL server service on both Primary and Secondary stations are on so that users can extract data from the Secondary for data analysis and report purposes Data replication 1s in real time as it 1s the Active Standby mode A built in backup utility tool is inc
40. ize 86848 Byte Remake Services and Controles sop 8 3 Port Reference Table The Port Reference Table lists some of the most commonly used ports with description as reference for users X Port Reference Table E d 5 xj Meru Echo Discard E 13 Close 17 Uniote 13 Chargen 21 Fie Protocol LL Default 20 Secure Shel SSH 23 Telnet 25 Sample Transfer Protocol SMTP 3 Time y New 3 who amp z Login Host Pretocal 53 Domain Hame Server DMS Wi Modify amp 3 File Transfer Protocol TF 79 Fieger w 20 werd wide web Bellen 33 Keiberos Y Authentication Past 2 Post Office Protocol v3 n i ar Find On the menu e Default will bring up the default listing e Modify allows changing of a port s description e New port definition can be created in New Click on the column title Port Number will sort the port numbers in ascending order and Description will sort the port descriptions in ascending alphabetical order 47
41. l primary dustered ayien Es uskrerem mid sccondery dualered ata Fs aer adhag due Ta clustering is iat iram male Ek fo orm PI FRG ca ee ee eg ee 3 3 Email Alert In two situations this feature can show importance 1 When a Failover occurs 2 When a Data Replication definition item is disabled by ClusterReplica process to prevent improper data replication With proper setup of this tool system administrators will be notified of happening of above situation To set it up from ClusterReplica s main status window click the tool bar button Options then select email alert setup Email Alert Configuration Email nobhcation vali be sent upon detection of cluster system Fallover and Iadback h Enable Emal Alert Serves Address 135 43 123 SMTP Part 25 Sender Email Address administratorteimpcompany com Recpient Email Addiess it john npoompary com Recipe Emad Address 2k manimecampary com Ie server requires sulberticalicey Accnunt Name administrator et aped m co 19 CHAPTER 4 Understand Failover and Failback Failover is the process in which the Secondary station of the cluster system take over the active role of the Primary station when failure of the Primary station is detected The failure of the Primary station can be either the entire system 1s down or one of the monitored
42. luded This tool assist users to backup data on the cluster server to a remote or local location for data protection Typically backup runs on the Secondary station so that the process will not interfere with network traffic serviced on the Primary station Besides database files ClusterReplica SQL 3 0 can also replicate other types of files including dynamically generated registry files ClusterReplica SQL 3 0 1s a powerful network server application of all purposes 1 2 Whom is ClusterReplica SQL 3 0 Designed For ClusterReplica SQL 3 0 is a fully functional data replication and Failover Failback server available software It is a data protection solution for small businesses and sub divisions of large corporations With automatic default configuration on MS SQL database file replication small business owners can utilize the software without depending on IT experts to setup and run the cluster server for their business needs When in its active active configuration SQL service on both Primary and Secondary stations are on so that users can extract data from the Secondary for data analysis and report purposes Data replication 1 in real time as it 1s the Active Standby mode For small businesses ClusterReplica SQL 3 0 comes with easy to use graphic user interface tools for software configuration Buttons lead to dialogue boxes for each of the feature setup The convenient online help with detailed instructions and the talk to the perso
43. mmend you use two network cards for the clustered systems one is dedicated to the internal communication one 1 dedicated to public connection This setting works for better security and efficiency in the cluster server 8 Click OK to move along the installation process If all goes well a few more clicks will bring you to the last screen Installshield Wizard InskaliShaeld Wizard Complete The Install held Wizard haz suecessiuls nstaled LlusterRepica SQL Edton YS Before vou can we Ihe program yau must nezter pow Computes oll ceshert my coempuler later Remove any disks from ther dives and then click Finish t complete setup Now you have successfully installed ClusterReplica SQL 3 0 software For Windows 2003 and XP systems system reboot at the end of installation of ClusterReplica SQL 3 0 is not required 10 2 3 Running the evaluation package ClusterReplica SQL 3 0 like all other Xlink s products has a 30 day evaluation period This means you can download a copy of ClusterReplica SQL 3 0 from our web site and test run it for 30 days to decide if the software fits your needs and system environment We encourage and recommend you to take advantage of the evaluation package and make sure you can use the software before purchase If after careful evaluation you decide to use the software a replace of the license key will allow you to continue using ClusterReplica SQL 3 0 without uninstall
44. n kind of technical support bring assurance to small business owners in using ClusterReplica The compatible pricing for the product also helps small businesses in their data protection decisions For sub division managers in a large corportation keeping the server available means to keep up the productivity Managers also need up to date data to make reports and better direct working strategies Small business owners and managers alike should all find it a compatitive and cost effective solustion for server availability and data protection ClusterReplica SQL 3 0 1s a useful tool to satisfy today s data protection need for all 1 3 How Does ClusterReplica SQL 3 0 Work ClusterReplica SQL binds two Microsoft Windows systems together to function as one cluster server so at any time there will be at least one system running to serve the network clients requests Within the cluster server one 1s taking the ACTIVE role to be the Primary station and the other taking the STANDBY role to be the Secondary station During standard condition both systems are running normally the Primary station will serve the network clients and replicate all newly changed data to the Secondary station to keep it current with all data This continuous data replication works on open and locked files including the database files The process ensures all changed data updated at all times No data will be lost due to system down or a service failure and server servic
45. nfiguration to replicate MS SQL database files on SQL server s default settings 6 1 Data Replication Setup Select Replication Area Setup from the main menu the configuration panel will look like following picture m E kebemepis PS SL dat eon xi Lee nip con PP un nea ant ET Hr SOL Mea SOL Seres 5 I Menhiro d A kream Pie Me SEL Seran a ET LCS A Feba D CoPi M SOL Seres M ES LSET ATA EO rege Pile Horeca Sere CAP coe Fie Mico SUL 9 erem OLS E d d PRESOLSERVER MES LSERVWER primary Clustered ayatem ia standalone amd arcesdany Catered system ia shuldiremn Dinin replication ia nol active duc to clustering ia nol in nirmal sedie Fer Help pran F Loca nier Prinaey Ful farre clu The stuff in each section of this panel are the files and services needed for ClusterReplica SQL 3 0 to correctly replicate all your SQL database files from the Primary station to the Secondary station at normal times and activate Failover process when a service or system failure 15 detected Other files and services to be replicated and monitored can be added by click the config button on the tool bar 36 Click Add by each section to add 1 File path for data replication 2 Dynamically created registry files 3 Servi
46. on for Keep alive time is 15 seconds This means at the end of each 15 seconds if no response from the remote system 15 detected by the local system local system will begin to operate in standalone mode The general rule of thumb 1s the faster the system s processing speed or busier the network traffic into the local system the longer the Keep alive time should be set Two things could happen when death is detected in a clustered system When the Primary station detected the Secondary station is dead data replication function will stop When the Secondary station detected the Primary station is dead it will assume the active role of the cluster server and function in standalone mode Communication port users can also change the communication port but be sure that the port number defined on both Primary and Secondary stations are the same when changes are made The next page is to setup IP numbers for internal communication of the cluster server Use the drop down menu to select the network card that has the IP number you want to use for this purpose Make sure the IP numbers on local and remote systems matches on the Primary and Secondary stations Step 2 Clusber Private Communication Addresses Sebup Step 2 Cluster Private Communication Addresses piyale communication addes k used by ClusterRiepkea bo resa en com micron amd dara on Both pama and zeacondaty ratrak car
47. responsibility to server the network clients This setting requires internal communication between the clustered servers as well as external communication between the server and the network clients 2 1 System Requirements for Running ClusterReplica SQL 3 0 Before install ClusterReplica SQL 3 0 please check the system requirements to make sure your system can handle the software properly e Two Microsoft Windows systems of 2000 XP 2003 platforms with ethernet connection e RAM 128MB or above e Hard disk 8GB or larger e Network connections between the clustered systems The most important condition being that the clustered systems must not be the domain controller or the DNS server Tips Using one or two network cards One network card If the systems you are going to install ClusterReplica SQL 3 0 each has one network card this means the cluster server operates with both internal and external communications moving within one subnet In this situation the cluster internal communication tangles up with the server clients communication The work load on the communication channel can be constantly overloaded There 1s also the concern of data processing speed and data transfer security Two network cards By installing two network cards on each of the clustered systems the internal and external communications are separated One set of IPs can then be devoted strictly to the internal communications of the clustered systems without t
48. software XLink specially designed an Error Report log file Users must follow listed steps to generate this report log From ClusterReplica user interface click Help on the menu bar then select Generate Report for Support 2 Following dialog box will come up Welcome to XLink Support Report Generator mmm This tool wall generate necessary setup information for Fise vend the generated report files to 1 Fill in the information with description of your problem then click Next Next screen confirms the input If you wish to make changes click Back to go back to previous page 2 Ifall looks right click finish to complete the process IEnter your information 1221231224 Repor Folders C vni CRRepon Now you have generated the error log Go to C XLinkCRReport to find the file Reportsetup txt Email this file to support xlink com and we will contact you for possible solutions If you have further questions regarding this product please feel free to contact us Phone 408 263 8201 fax 408 263 8203 email support xlink com NOTE Please always include the product name in the subject line when you send email for support Thanks CHAPTER 2 Installing Removing ClusterReplica SQL 3 0 A cluster server binds two or more computer systems together to function as one The clustered systems need to communicate with each other At the same time one of the clustered systems will take up the
49. ster Failover using ClusterReplica SQL 3 0 are The clustered systems are in same network segment 2 The clustered systems are in different network segment The most important condition being that the clustered systems must not be the domain controller or the DNS server The Failover setup considerations are summed up in following When both Primary and Secondary stations are in the same network segment In this system environment three situations can be considered 1 Local host table or IP address is used the Cluster IP must be set MS Netbios is used to resolve network address the Cluster Name must be set 3 DNS is configured the Cluster Name and Cluster IP can be optional In the case where all three situations listed above are involved users need to setup the Cluster Name Cluster IP and the DNS configuration in Failover setup When the Primary and Secondary stations are in different network subnet In this system environment only DNS configuration can be used for Failover setup Enable Cluster Name or Cluster IP may result failure of cluster Failover 30 5 22 Configure Failover Setup From the main menu select Failover Setup to bring up the configuration panel To begin configuration process click Modify TS SOL Edition 2 0 Chester D zablad Fleplicicn poke Loca ase p prins elation Faki led Sie
50. t using the Failover amp nai 22 4 2 Pailback or Not Pallback uoo eh ted uv ao adis 22 4 2 ClusterReplica SQL 3 0 offers two ways to Failback 22 4 22 Advantage and disadvantages in each of the Failback method 24 P P 26 Cluster Configuration 26 Sob RUCNO EU 26 gt eed 30 5 21 Network Configuration Concerning 30 3 22 Failover Setups ec dota 31 FeO ccsaisvaad dition 36 Replication Area Configuration 36 Data FR EpliCaviOMm 36 6 2 Replication Advanced Seti oos cce Rep etis ius 38 6 3 Replication Verification Schedule reete tentoria 39 6 4 Compare replicatton TIG TREO eios eee Dro Fees ato 40 Data se aot utate tette endo 4 bd EN 42 Scheduled File Backup eem 42 7 1 Two examples on setup backup Error Bookmark not defined CHAPTER 8 50 IECUR EU 45 Otner LOOMS sexisesixiniesitexbnbrliaivedasssadirtasmiwss aa d 45 Sl cat ten tut I E 45 8 2 Network Applicat
51. te 2 When Cluster IP in Failover setup step 2 1s enabled the network The first step in enable DNS setup 1s to check the box Enable DNS Failover Redirection Following is an example on DNS configuration card that is binding the Cluster IP must check off Register this connection s addresses in DNS in the Advanced TCP IP settings This must be done on both Primary and Secondary stations 34 Step 4 ONS Redirection Setup Enable DNS redeechon Finan OHS Server 15 2 14 dd Secondup DHS Semer O 0 Oo 0o Full quale DNS nami rest ansemat Local IP hor sediechon 15 2 11 1 1S 2 BEERS Here is the summary page to show you the complete setup If there is anything you want to change click Back to make the changes Otherwise click Finish to complete the process Chuster Failover Setup Summary cer IF 19521 25 255 255 255 0 te Family Fact HIC BEEN zl NI 35 CHAPTER 6 Replication Area Configuration The file replication function of ClusterReplica SQL 3 0 is among the most powerful and efficient in its class Besides handling all types of files including SQL database files and some registry files ClusterReplica SQL 3 0 replicates files in real time even when the file is open and being used With successful installation ClusterReplica SQL 3 0 is set by default in Replication only mode with automatic co
52. the ican ta gai a general dazc Herc ai Pere T22121 0216 z The Replication Status Window at the bottom of the screen indicating the initial file sync is in progress It is recommended that no user action is to be taken before the completion of this initial file sync process 12 When the initial file sync 1s completed the Replication Status Window will show as in next picture Now you can proceed to add additional file replication paths and application services for failover monitoring Yr PTS SHE ifi en 3 0 ele Setup oc Sext Vies CLRH Log Cluster Status Cluster slatus vire allows uror dp zumant clustered strate on ihe syam ican bo gal gengral deaciplibns SwkhRue 2 x Anke Pr nae S prins si Loc y n Gha D 122 1212110 oad EELS 11 53 10 Syrum E Ferera 4 Control 12112212116 a Gras Tira D AELZIOS 1447 BEL Chute sinus T Schede Oita Veiis T I Feed ge Fiar rece For Help mran F1 The detailed status of data replication and file backup can be checked by looking into their log files Click View on the tool bar and view each of the status logs as shown in following pictures Cluster Status Log Im Cluster Status L
53. ve the most current data In our example Failback with data recovery from remote system is selected Failback data option me p NM Failback Data Option Local System rs selected lo be primary Falback wilh data recovery wil replace local date in liste vath he ores That curently on remobe satem Please refer to Fallback of the Cluster System rr Help Falback eslhout data recevery from remote system Falback vathout data recess will ignore any changes or updates of data repkcalion lists on remote system and 1eplace remate dete contents with current local qesoureEE Choose this eplien only d dala on remote system can be discarded and replaced with bacal data 23 e The next picture shows the Summary page which tells you that you have choose to let the newly start server to become the Primary station and have the current data on the active standalone station will be copied over If this is not what you really wanted click on Back to go to the previous pages to make the changes Clustering role confirmation summary E x Summary Local cluster member Primary Station Actions after clicking on Finish 1 Local system wal become secondary station frst 2 Remote will stay az primary station 3 Data on remote system wil be replicated bo bocal apatan 4 Upon compaction of Fal replication koca sysher till resume the role of pnma cluster
54. y the keys for specific applications Monitored Service All files that need specific application to access for example the database files require SQL application to access would require to have the application service to be added in this listing so that when failover takes place these services will be activated on the Secondary station These services are also monitored for service level failover which will happen when any one of the services in the list is detected of malfunction 37 6 2 Replication Advanced Settings ClusterReplica SQL 3 0 replicates modified data of database files and non database files in the point in time fashion With successful installation some default values of memory space allocation and timing arrangement for data replication operation are set as recommended values for best performance Users however still have the power to change these values to best fit their system environment This flexibility promises the highest product efficiency and user satisfaction Advanced setting m UU N R E E x Virtual memor usage Minimum virtual memory used for intemal cache p Maximum virbual memory used for internal cache 292 AB CEOGOG NOGDEGNGUE N GOE 420222 lt lt a E a lt Replication versioan point Version point timeout is used to generate safe dal
Download Pdf Manuals
Related Search