6.3
Table Of Contents
- Administrator Guide
- Contents
- About This Book
- Getting Started
- Introduction
- Configuring vCenter Server Heartbeat
- Server Configuration Wizard
- Configuring the Machine Identity
- Configuring the Server Role
- Configuring the Client Connection Port
- Configuring Channel Routing
- Configuring the Default Channel Port
- Configuring Low Bandwidth Module
- Configuring Public IP Addressing
- Configuring for Identical Nodes
- Configuring for Non-Identical Nodes
- Managing vCenter Server Heartbeat License Keys
- Configuring Message Queue Logs
- Configuring the Maximum Disk Usage
- System Administration and Management
- Server Protection
- Network Protection
- Application Protection
- Status and Control
- vCenter Server Heartbeat Console
- About vCenter Server Heartbeat Console
- Work with Groups and Pairs
- Add, Edit, Move, and Remove Pairs in VCenter Server Heartbeat Groups
- Review the Status of vCenter Server Heartbeat Groups and Pairs
- Exit vCenter Server Heartbeat Console
- Shut Down Windows Without Stopping vCenter Server Heartbeat
- Performance Protection
- Data Protection
- Data Protection Overview
- Other Administrative Tasks
- Troubleshooting
- Troubleshooting Unexpected Behaviors
- Two Active Servers
- Two Passive Servers
- Synchronization Failures
- Registry Status is Out of Sync
- Channel Drops
- Subnet or Routing Issues
- MaxDiskUsage Errors
- MaxDiskUsage Error Messages
- [L9]Exceeded the Maximum Disk Usage (VCChannelExceededMaxDiskUsageException)
- [L9]Exceeded the Maximum Disk Usage on the ACTIVE Server
- [L9]Exceeded the Maximum Disk Usage on the PASSIVE Server
- [L20]Out of Disk Space (VCChannelOutOfDiskSpaceException)
- Application Slowdown
- Poor Application Performance
- Both Servers Can Accommodate the Initial Load but the Load Has Increased
- One Server Can Provide Adequate Resource Support, but the Other Cannot
- Scheduled Resource Intensive Tasks
- Glossary
VMware, Inc. 35
Chapter 3 Server Protection
ThisisparticularlyusefulwhenmaintenanceworkperformedonthePrimaryserverrequiresrebootingthe
server.
Sinceamanagedswitchovercannotbeperformedduringsynchronization,itisimportanttoreviewthequeue
informationpriortoattemptingamanagedswitchover.Ifthequeuesarelarge,fileoperationsontheactive
serverare
highandforthisreasonitmaybeprudenttodelayamanagedswitchoverduetothelengthoftime
requiredtocompletelyclearthequeue.QueuelengthscanbeviewedintheData:Traffic/Queuespageofthe
vCenterServerHeartbeatConsole.
PriortoperformingworkonthePrimaryserver,a
managedswitchovercanbetriggeredbyselectingthe
SecondaryserverandclickingMakeActiveontheServer:Summarypage.Thischangestheserverrolessuch
thattheactiveserverbecomespassiveandthepassiveserverbecomesactive.Thismeansusersareableto
workcontinuouslywhilethePrimaryserverisoff
line.
WhenthePrimaryserverisbackupandrunning,themanagedswitchovercanbetriggeredagainsothatthe
Primaryserverbecomesactiveandthepreviouslyactiveserverbecomespassive.
Recovering From a Failover
Afailoverdiffersfromaswitchover.Aswitchoverisacontrolledswitch(initiatedmanuallyfromthevCenter
ServerHeartbeatConsole,orinitiatedbyvCenterServerHeartbeatwhenpreconfigured)betweentheactive
andpassiveservers.Afailoverhappenswhenanyofthefollowingfailontheactiveserver:power,hardware,
orVMware
Channelcommunications.Thepassiveserverwaitsapreconfiguredtimeafterthefirstmissed
heartbeatbeforeinitiatingafailover.Whenthisperiodexpires,thepassiveserverautomaticallyassumesthe
activeroleandstartstheprotectedapplications.
ThefollowingrecoveryscenarioisbasedonvCenterServerHeartbeatconfigurationwiththePrimaryserver
as
activeandtheSecondaryserveraspassive.
AfailoverhasoccurredandtheSecondaryserverisnowrunningastheactiveserver.
1Revieweventlogsonbothserverstodeterminethecauseofthefailover.Forassistance,usetheLog
CollectortooltocollectinformationandsendtheoutputtoVMware
Support.
2IfanyofthefollowingissuesexistonthePrimaryserver,performingaswitchbacktothePrimaryserver
maynotbepossibleuntilotherimportantactionsarecarriedout.DonotrestartvCenterServerHeartbeat
untilthefollowingissueshavebeenresolved:
HardDiskFailure–Replacethedefectiveharddisk.
PowerFailure–RestorepowertothePrimaryserver.
Virus–Cleantheserverofallviruses.
Communications–Replaceorrepairthephysicalnetworkhardware.
BlueScreen–Determinecauseandresolve.Asrequired,submitthedumpfiletoVMwareSupport
(www.vmware.com/support)foranalysis.
3RuntheConfigureServerwizardandverifytheserveridentityissettoPrimaryandtheroleissetto
passive.ClickFinishtoacceptthechanges.
4DisconnecttheVMwareChannelnetwork
cablesordisablethenetworkcard.
5Resolvethelistofpossiblefailures.
6Restartthisserverandreconnectorenablethenetworkcardagain.
I
MPORTANTThemanagedswitchoverprocessmaybeperformedatanytimeaslongasthesystemsarefully
synchronizedwithrespecttodatafilesandregistryreplication.Switchoverscannotbeperformedifeitherserveris
inanunsynchronizedorunknownstate.
NOTEWhenfailoverconditions,suchasapowerfailure,causefailuresinbothactiveandpassiveservers,a
conditionmayresultthatcausesbothserverstorestartinpassivemode.Inthissituation,manualintervention
isrequired.SeeAppendix 10,“TwoPassiveServers,”onpage 67.