OracleDBAInterviewQuestionsandAnswersRAC
OracleRACInterviewQuestionsandAnswers
HowdoesOCSSDstartsfirstifvotingdisk&OCRresidesinASMDiskgroups?
YoumightwonderhowCSSD,whichisrequiredtostarttheclusteredASMinstance,canbestartedif
votingdisksarestoredinASM?
Thissoundslikeachickenandeggproblem:
withoutaccesstothevotingdisksthereisnoCSS,hencethenodecannotjointhecluster.
Butwithoutbeingpartofthecluster,CSSDcannotstarttheASMinstance.
TosolvethisproblemtheASMdiskheadershavenewmetadatain11.2:
youcanusekfedtoreadtheheaderofanASMdiskcontainingavotingdisk.
[Link]
instancetobeup.
Oncethevotingdisksarelocated,CSScanaccessthemandjoinsthecluster.
WhatisgsdctlinRAC?listgsdctlcommandsinOracleRAC?
GSDCTLstandsforGlobalServiceDaemonControl,wecanusegsdctlcommandstostart,stop,andobtainthe
statusoftheGSDserviceonanyplatform.
Theoptionsforgsdctlare:
$gsdctlstartTostarttheGSDservice
$gsdctlstopTostoptheGSDservice
$gsdctlstatToobtainthestatusoftheGSDservice
Logfilelocationforgsdctl:
$ORACLE_HOME/srvm/log/gsdaemon_node_name.log
WhatisRAC?
RACstandsforRealApplicationcluster.
ItisaclusteringsolutionfromOracleCorporationthatensureshighavailabilityofdatabasesbyproviding
instancefailover,mediafailoverfeatures.
OracleRACisaclusterdatabasewithasharedcachearchitecturethatovercomesthelimitationsoftraditional
sharednothingandshareddiskapproachestoprovideahighlyscalableandavailabledatabasesolutionforall
thebusinessapplications.
OracleRACprovidesthefoundationforenterprisegridcomputing.
WhatisOracleRACOneNode?
OracleRAConeNodeisasingleinstancerunningononenodeoftheclusterwhilethe2ndnodeisincold
[Link]
[Link]
thisfeatureisthatitprovidesacoldfailoversolutionanditautomatestheinstancerelocationwithoutany
downtimeanddoesnotneedamanualintervention.Oracleintroducedthisfeaturewiththereleaseof11gR2
(availablewithEnterpriseEdition).
WhatisRACandhowisitdifferentfromnonRACdatabases?
OracleRealApplicationclustersallowsmultipleinstancestoaccessasingledatabase,theinstanceswillbe
runningonmultiplenodes.
InRealApplicationClustersenvironments,allnodesconcurrentlyexecutetransactionsagainstthesame
database.
RealApplicationClusterscoordinateseachnode'saccesstotheshareddatatoprovideconsistencyand
integrity.
WhataretheadvantagesofRAC(RealApplicationClusters)?
Reliabilityifonenodefails,thedatabasewon'tfail
Availabilitynodescanbeaddedorreplacedwithouthavingtoshutdownthedatabase
Scalabilitymorenodescanbeaddedtotheclusterastheworkloadincreases
WhatisOracleRACOneNode?
OracleRAConeNodeisasingleinstancerunningononenodeoftheclusterwhilethe2ndnodeisincold
[Link]
[Link]
thisfeatureisthatitprovidesacoldfailoversolutionanditautomatestheinstancerelocationwithoutany
downtimeanddoesnotneedamanualintervention.Oracleintroducedthisfeaturewiththereleaseof11gR2
(availablewithEnterpriseEdition).
WhatisCacheFusion?
[Link]
withintheclusterandanotherinstanceisinneedofthesameblock,itiseasytogettheblockimagefromthe
[Link]
[Link](GES)monitorsand
Instanceenqueueprocessmanagesthecachefusion.
WhatcommandwouldyouusetochecktheavailabilityoftheRACsystem?
crs_stattv(tvareoptional)
HowdoweverifythatRACinstancesarerunning?
SQL>select*fromV$ACTIVE_INSTANCES
ThequerygivestheinstancenumberunderINST_NUMBERcolumn,host_:instancenameunderINST_NAME
column.
HowcanyouconnecttoaspecificnodeinaRACenvironment?
tnsnames.oraensurethatyouhaveINSTANCE_NAMEspecifiedinit.
Whichisthe"MASTERNODE"inRAC?
Thenodewiththelowestnodenumberwillbecomemasternodeanddynamicremasteringoftheresourceswill
takeplace.
Tofindoutthemasternodeforparticularresource,youcanqueryv$ges_resourceforMASTER_NODE
column.
Tofindoutwhichisthemasternode,[Link]"masternodenumber".
whenthefirstmasternodefailsintheclusterthelowestnodenumberwillbecomemasternode.
WhatcomponentsinRACmustresideinsharedstorage?
Alldatafiles,controlfiles,SPFIles,redologfilesmustresideonclusterawareshredstorage.
Givefewexamplesforsolutionsthatsupportclusterstorage?
ASM(automaticstoragemanagement),
Rawdiskdevices,
Networkfilesystem(NFS),
OCFS2and
OCFS(OracleClusterFiesystems).
WhatareOracleClusterComponents?
[Link](HAIP)
[Link](OCR/VotingDisk)
[Link]
[Link]
WhatareOracleRACComponents?
VIP,Nodeappsetc.
WhatareOracleKernelComponents?
BasicallyOraclekernelneedtoswitchedonwithRACOnoptionwhenyouconverttoRAC,thatisthedifference
asitfacilitatesfewRACbgprocesslikeLMON,LCK,LMD,LMSetc.
HowtoturnonRAC?
#linktheoraclelibraries
$cd$ORACLE_HOME/rdbms/lib
$makefins_rdbms.mkrac_on
#rebuildoracle
$cd$ORACLE_HOME/bin
$relinkoracle
DiskarchitechtureinRAC?
SAN(StorageAreaNetworks)generallyusingfibretoconnecttotheSAN
NAS(NetworkAttachedStorage)generallyusinganetworktoconnecttotheNASusingeitherNFS,ISCSI
WhatisOracleClusterware?
TheClusterwaresoftwareallowsnodestocommunicatewitheachotherandformstheclusterthatmakesthe
nodesworkasasinglelogicalserver.
ThesoftwareisrunbytheClusterReadyServices(CRS)usingtheOracleClusterRegistry(OCR)thatrecords
andmaintainstheclusterandnodemembershipinformationandthevotingdiskwhichactsasatiebreaker
[Link]
diskwhentheclusterisrunning.
RealApplicationClusters
OracleRACisaclusterdatabasewithasharedcachearchitecturethatovercomesthelimitationsoftraditional
sharednothingandshareddiskapproachestoprovideahighlyscalableandavailabledatabasesolutionforall
[Link].
OraclesRealApplicationClusters(RAC)optionsupportsthetransparentdeploymentofasingledatabase
acrossaclusterofservers,[Link]
runningonclustersprovidesOracleshighestlevelofcapabilityintermsofavailability,scalability,andlowcost
computing.
OneDBopenedbymultipeinstancessothethedbllbeHighlyAvailableifaninstancecrashes.
[Link]
clustersupportandalloweachnodetoknowwhichnodesbelongtotheclusterandareavailableandwith
OracleCluterwaretoknowwhichnodeshavefailedandtoejectthenfromthecluster,sothaterrorsonthat
nodecanbecleared.
OracleClusterwarehastwokeycomponentsClusterRegistryOCRandVotingDisk.
Theclusterregistryholdsallinformationaboutnodes,instances,servicesandASMstorageifused,italso
containsstateinformationietheyareavailableanduporsimilar.
Thevotingdiskisusedtodetermineifanodehasfailed,[Link]
deemedtonolongerbelongtothemajoritythenitisforciblyrebootedandwillaftertherebootadditselfagain
thethesurvivingclusternodes.
WhataretheOracleClusterwarekeycomponents?
OracleClusterwarehastwokeycomponentsClusterRegistryOCRandVotingDisk.
WhatisVotingDiskandOCR?
VotingDisk
OracleRACusesthevotingdisktomanageclustermembershipbywayofahealthcheckandarbitratescluster
[Link].
Anodemustbeabletoaccessmorethanhalfofthevotingdisksatanytime.
Forexample,ifyouhave3votingdisksconfigured,thenanodemustbeabletoaccessatleasttwoofthe
[Link],or
removed,fromthecluster.
OracleClusterRegistry(OCR)
Theclusterregistryholdsallinformationaboutnodes,instances,servicesandASMstorageifused,italso
containsstateinformationietheyareavailableanduporsimilar.
TheOCRmustresideonshareddiskthatisaccessiblebyallofthenodesinyourcluster.
Whataretheadministrativetasksinvolvedwithvotingdisk?
Followingadministrativetasksareperformedwiththevotingdisk:
1)Backingupvotingdisks
2)RecoveringVotingdisks
3)Addingvotingdisks
4)Deletingvotingdisks
5)Movingvotingdisks
Canyouaddvotingdiskonline?Doyouneedvotingdiskbackup?
Yes,asperdocumentation,ifyouhavemultiplevotingdiskyoucanaddonline,butifyouhaveonlyonevoting
disk,bythatclusterwillbedownasitslostyoujustneedtostartcrsinexclusivemodeandaddthevotedisk
using
crsctladdvotedisk<path>
WhatistheOracleRecommendationforbackingupvotingdisk?
Oraclerecommendsustousetheddcommandtobackupthevotingdiskwithaminimumblocksizeof4KB.
Howdowebackupvotingdisks?
1)Oraclerecommendsthatyoubackupyourvotingdiskaftertheinitialclustercreationandafterwecomplete
anynodeadditionordeletionprocedures.
2)First,asrootuser,stopOracleClusterware(withthecrsctlstopcrscommand)[Link],determine
thecurrentvotingdiskbyissuingthefollowingcommand:
crsctlqueryvotediskcss
3)Then,issuetheddorocopycommandtobackupavotingdisk,asappropriate.
Givethesyntaxofbackingupvotingdisks:
OnLinuxorUNIXsystems:
ddif=voting_disk_nameof=backup_file_name
where,
voting_disk_nameisthenameoftheactivevotingdisk
backup_file_nameisthenameofthefiletowhichwewanttobackupthevotingdiskcontents
OnWindowssystems,usetheocopycommand:
copyvoting_disk_namebackup_file_name
HowdoweverifyanexistingcurrentbackupofOCR?
WecanverifythecurrentbackupofOCRusingthefollowingcommand:ocrconfigshowbackup
YouhavelostOCRdisk,whatisyournextstep?
Theclusterstackwillbedownduetothefactthatcssdisunabletomaintaintheintegrity,thisistruein10g,
From11gR2onwards,thecrsdstackwillbedown,[Link]
restoringtheautomaticbackuporimportthemanualbackup,
WhatarethemajorRACwaitevents?
InaRACenvironmentthebuffercacheisglobalacrossallinstancesintheclusterandhencetheprocessing
[Link]
GCCRrequest:thetimeittakestoretrievethedatafromtheremotecache
Reason:RACTrafficUsingSlowConnectionorInefficientqueries(poorlytunedquerieswillincreasethe
[Link]
oftenablockwillneedtobereadfromaremoteinstanceviatheinterconnect.)
GCBUFFERBUSY:Itisthetimetheremoteinstancelocallyspendsaccessingtherequesteddatablock.
WhatdoyoudoifyouseeGCCRBLOCKLOSTintop5TimedEventsinAWRReport?
Thisismostlikelyduetoafaultininterconnectnetwork.
Checknetstats
ifyousee"fragmentsdropped"or"packetreassembliesfailed",Workwithyoursystemadministratorfindthe
faultwithnetwork.
Howdoyoutroubleshootnodereboot?
Pleasecheckmetalink...
Note265769.1TroubleshootingCRSReboots
Note.559365.1UsingDiagwaitasadiagnostictogetmoreinformationfordiagnosingOracleClusterwareNode
evictions.
Srvctlcannotstartinstance,IgetthefollowingerrorPRKP1001CRS0215,howeversqlpluscanstartit
onbothnodes?Howdoyouidentifytheproblem?
SettheenvironmentalvariableSRVM_TRACEtotrue..[Link]
detailederrorstack.
WhatareOracleClusterwareprocessesfor10gonUnixandLinux?
ClusterSynchronizationServices(ocssd)Managesclusternodemembershipandrunsastheoracle
userfailureofthisprocessresultsinclusterrestart.
ClusterReadyServices(crsd)Thecrsprocessmanagesclusterresources(whichcouldbeadatabase,an
instance,aservice,aListener,avirtualIP(VIP)address,anapplicationprocess,andsoon)basedonthe
resource'[Link],stop,monitorandfailover
[Link]
Eventmanagerdaemon(evmd)Abackgroundprocessthatpublisheseventsthatcrscreates.
ProcessMonitorDaemon(OPROCD)ThisprocessmonitortheclusterandprovideI/[Link]
performsitscheck,stopsrunning,andifthewakeupisbeyondtheexpectedtime,thenOPROCDresetsthe
[Link].
OPROCDusesthehangchecktimeronLinuxplatforms.
RACG(racgmain,racgimon)ExtendsclusterwaretosupportOraclespecificrequirementsandcomplex
[Link].
WhatareOracledatabasebackgroundprocessesspecifictoRAC?
[Link]
[Link]
GES(GlobalEnqueueService),GCS(GlobalCacheService)[Link]
composedoffollowingbackgroundprocesses:
ACMSAtomicControlfiletoMemoryService(ACMS)
GTX0jGlobalTransactionProcess
LMONGlobalEnqueueServiceMonitor
LMDGlobalEnqueueServiceDaemon
LMSGlobalCacheServiceProcess
LCK0InstanceEnqueueProcess
RMSnOracleRACManagementProcesses(RMSn)
RSMNRemoteSlaveMonitor
ToensurethateachOracleRACdatabaseinstanceobtainstheblockthatitneedstosatisfyaqueryor
transaction,OracleRACinstancesusetwoprocesses,theGlobalCacheService(GCS)andtheGlobal
EnqueueService(GES).TheGCSandGESmaintainrecordsofthestatusesofeachdatafileandeachcached
blockusingaGlobalResourceDirectory(GRD).TheGRDcontentsaredistributedacrossalloftheactive
instances.
WhatisGRD?
[Link]
[Link]
helpsindataintegrity.
WhatisACMS?
[Link]
ensuresadistributedSGAmemoryupdate(ie)SGAupdatesaregloballycommittedonsuccessorglobally
abortedineventofafailure.
WhatisSCANlistener?
Ascanlistenerissomethingthatadditionaltonodelistenerwhichlistenstheincomingdbconnectionrequests
fromtheclientwhichgotthroughthescanIP,itgotendpointsconfiguredtonodelistenerwhereitroutesthedb
connectionrequeststoparticularnodelistener.
[Link].
Enabling/disablingSCANIPismostlyusedinoracleappsenvironmentbytheconcurrentmanager(kindofjob
schedulerinoracleapps).
StepstodisabletheSCANIP,
[Link].
[Link]
srvctlstopscan_listener
[Link]
srvctlstopscan(thiswillstopthescanvip's)
[Link]
srvctldisablescan
Whatarethedifferentnetworkcomponentsarein10gRAC?
public,private,andvipcomponents
Privateinterfacesisforintranodecommunication.
[Link]
node,thisisthereasonthatallapplicationsshouldbasedonvipcomponentsmeanstnsentriesshouldhavevip
entryinthehostlist
Whatisaninterconnectnetwork?
[Link]
networkusesaswitch/multipleswitchesthatonlythenodesintheclustercanaccess.
Whatistheuseofclusterinterconnect?
ClusterinterconnectisusedbytheCachefusionforinterinstancecommunication.
Howcanweconfiguretheclusterinterconnect?
ConfigureUserDatagramProtocol(UDP)onGigabitEthernetforclusterinterconnects.
OnUNIXandLinuxsystemsweuseUDPandRDS(Reliabledatasocket)protocolstobeusedbyOracle
Clusterware.
WindowsclustersusetheTCPprotocol.
WhatisthepurposeofPrivateInterconnect?
Clusterwareusestheprivateinterconnectforclustersynchronization(networkheartbeat)anddaemon
[Link].
RACusestheinterconnectforcachefusion(UDP)andinterprocesscommunication(TCP).CacheFusionis
theremotememorymappingofOraclebuffers,sharedbetweenthecachesofparticipatingnodesinthecluster.
WhatisavirtualIPaddressorVIP?
AvirtualIPaddressorVIPisanalternateIPaddressthattheclientconnectionsuseinsteadofthestandard
[Link],weneedtoreserveaspareIPaddressforeachnode,andtheIP
addressesmustusethesamesubnetasthepublicnetwork.
WhatistheuseofVIP?
Ifanodefails,thenthenode'sVIPaddressfailsovertoanothernodeonwhichtheVIPaddresscanacceptTCP
connectionsbutitcannotacceptOracleconnections.
WhydowehaveaVirtualIP(VIP)inOracleRAC?
WithoutusingVIPsorFAN,clientsconnectedtoanodethatdiedwilloftenwaitforaTCPtimeoutperiod(which
canbeupto10min)[Link],youdon'treallyhaveagoodHAsolutionwithoutusing
VIPs.
Whenanodefails,theVIPassociatedwithitisautomaticallyfailedovertosomeothernodeandnewnodere
[Link]
node,[Link]
immediately.
GivesituationsunderwhichVIPaddressfailoverhappens?
VIPaddressesfailoverhappenswhenthenodeonwhichtheVIPaddressrunsfailsallinterfacesfortheVIP
addressfails,allinterfacesfortheVIPaddressaredisconnectedfromthenetwork.
WhatisthesignificanceofVIPaddressfailover?
WhenaVIPaddressfailoverhappens,ClientsthatattempttoconnecttotheVIPaddressreceivearapid
[Link]'thavetowaitforTCPconnectiontimeoutmessages.
WhatistheuseofaserviceinOracleRACenvironment?
[Link]
rulesandcharacteristicstocontrolhowusersandapplicationsconnecttodatabaseinstances.
WhatarethecharacteristicscontrolledbyOracleservicesfeature?
Thecharacteristicsincludeauniquename,workloadbalancing,failoveroptions,andhighavailability.
WhatenablestheloadbalancingofapplicationsinRAC?
OracleNetServicesenabletheloadbalancingofapplicationconnectionsacrossalloftheinstancesinan
OracleRACdatabase.
Whatarethetypesofconnectionloadbalancing?
ConnectionWorkloadmanagementisoneofthekeyaspectswhenyouhaveRACinstancesasyouwantto
distributetheconnectionstospecificnodes/instanceorthosehavelessload.
Therearetwotypesofconnectionloadbalancing:
[Link](alsocalledasconnecttimeloadbalancing)
[Link](alsocalledasListenerconnectionloadbalancing)
Whatisthedifferencebetweenserversideandclientsideconnectionloadbalancing?
[Link]
loadbalancinglistenerusesaloadbalancingadvisorytoredirectconnectionstotheinstanceprovidingbest
service.
ClientSideloadbalancing:Oracleclientsideloadbalancingfeatureenablesclientstorandomizethe
connectionrequestsamongalltheavailablelistenersbasedontheirload.
Antnsentrythatcontainsallnodesentriesanduseload_balance=on(defaultitson)willusetheconnecttime
loadbalancingorclientsideloadbalancing.
SampleClientSideTNSEntry:
finance=
(DESCRIPTION=
(ADDRESS=(PROTOCOL=TCP)(HOST=myrac2vip)(PORT=2042))
(ADDRESS=(PROTOCOL=TCP)(HOST=myrac1vip)(PORT=2042))
(ADDRESS=(PROTOCOL=TCP)(HOST=myrac3vip)(PORT=2042))
(LOAD_BALANCE=yes)
(CONNECT_DATA=
(SERVER=DEDICATED)
(SERVICE_NAME=FINANCE)(FAILOVER=ON)
(FAILOVER_MODE=(TYPE=SELECT)(METHOD=BASIC)(RETRIES=180)(DELAY=5))
)
)
Serversideloadbalancing:Thisimprovestheconnectionperformancebybalancingthenumberofactive
[Link](sharedservers),the
[Link],PMONis
awareofallinstancesloadanddispatchers,anddependingontheloadinformationPMONredirectsthe
connectiontotheleastloadednode.
InaRACenvironment,*.remote_listenerparameterwhichisatnsentrycontainingallnodesaddressesneedto
settoenabletheloadbalanceadvisoryupdatestoPMON.
SampleTnsentryshouldbeinaninstancesofRACcluster,
local_listener=LISTENER_MYRAC1
remote_listener=LISTENERS_MYRACDB
WhataretheadministrativetoolsusedforOracleRACenvironments?
OracleRACclustercanbeadministeredasasingleimageusingthebelow
OEM(EnterpriseManager),
SQL*PLUS,
Servercontrol(SRVCTL),
ClusterVerificationUtility(CLUVFY),
DBCA,
NETCA
NamesomeOracleClusterwaretoolsandtheiruses?
OIFCFGallocatinganddeallocatingnetworkinterfaces.
OCRCONFIGCommandlinetoolformanagingOracleClusterRegistry.
OCRDUMPIdentifytheinterconnectbeingused.
CVUClusterverificationutilitytogetstatusofCRSresources.
WhatisthedifferencebetweenCRSCTLandSRVCTL?
crsctlmanagesclusterwarerelatedoperations:
StartingandstoppingOracleClusterware
EnablinganddisablingOracleClusterwaredaemons
Registeringclusterresources
srvctlmanagesOracleresourcerelatedoperations:
Startingandstoppingdatabaseinstancesandservices
Alsofrom11gR2managestheclusterresourceslikenetwork,vip,disksetc
HowdoweremoveASMfromaOracleRACenvironment?
[Link]
removedusingsrvctltoolasfollows:
srvctlstopasmnnode_name
srvctlremoveasmnnode_name
WecanverifyifASMhasbeenremovedbyissuingthefollowingcommand:
srvctlconfigasmnnode_name
HowdoweverifythataninstancehasbeenremovedfromOCRafterdeletinganinstance?
Issuethefollowingsrvctlcommand:
srvctlconfigdatabaseddatabase_name
cdCRS_HOME/bin
./crs_stat
WhatarethemodesofdeletinginstancesfromORacleRealApplicationclusterDatabases?
WecandeleteinstancesusingsilentmodeorinteractivemodeusingDBCA(DatabaseConfigurationAssistant).
Whatarethebackgroundprocessthatexistsin11gr2andfunctionality?
ProcessNameFunctionality
crsdTheCRSdaemon(crsd)managesclusterresourcesbasedonconfigurationinformationthatisstored
inOracleClusterRegistry(OCR)[Link],stop,monitor,andfailoveroperations.
Thecrsdprocessgenerateseventswhenthestatusofaresourcechanges.
cssdClusterSynchronizationService(CSS):Managestheclusterconfigurationbycontrollingwhichnodes
[Link]
certifiedthirdpartyclusterware,thenCSSprocessesinterfaceswithyourclusterwaretomanagenode
[Link]:theCSSdaemon(ocssd),theCSSAgent
(cssdagent),andtheCSSMonitor(cssdmonitor).Thecssdagentprocessmonitorstheclusterandprovides
input/[Link](oprocd),also
[Link].
diskmonDiskMonitordaemon(diskmon):Monitorsandperformsinput/outputfencingforOracleExadata
[Link],thediskmon
daemonisalwaysstartedwhenocssdisstarted.
evmdEventManager(EVM):IsabackgroundprocessthatpublishesOracleClusterwareevents
mdnsdMulticastdomainnameservice(mDNS):[Link]
processonLinuxandUNIX,andaserviceonWindows.
gnsdOracleGridNamingService(GNS):IsagatewaybetweentheclustermDNSandexternalDNS
[Link].
onsOracleNotificationService(ONS):IsapublishandsubscribeserviceforcommunicatingFast
ApplicationNotification(FAN)events
oraagentoraagent:[Link]
[Link]
11gRelease1(11.1).
orarootagentOraclerootagent(orarootagent):IsaspecializedoraagentprocessthathelpsCRSDmanage
resourcesownedbyroot,suchasthenetwork,andtheGridvirtualIPaddress
oclskdClusterkilldaemon(oclskd):Handlesinstance/nodeevictionsrequeststhathavebeenescalatedto
CSS
gipcdGridIPCdaemon(gipcd):Isahelperdaemonforthecommunicationsinfrastructure
ctssdClustertimesynchronisationdaemon(ctssd)tomanagethetimesyncrhonizationbetweennodes,
ratherdependingonNTP
Underwhichuserorownertheprocesswillstart?
ComponentNameoftheProcessOwner
OracleHighAvailabilityServiceohasdinit,root
ClusterReadyService(CRS)ClusterReadyServicesroot
ClusterSynchronizationService(CSS)ocssd,cssdmonitor,cssdagentgridowner
EventManager(EVM)evmd,evmloggergridowner
ClusterTimeSynchronizationService(CTSS)octssdroot
OracleNotificationService(ONS)ons,eonsgridowner
OracleAgentoragentgridowner
OracleRootAgentorarootagentroot
GridNamingService(GNS)gnsdroot
GridPlugandPlay(GPnP)gpnpdgridowner
Multicastdomainnameservice(mDNS)mdnsdgridowner
Whatisthemajordifferencebetween10gand11gRAC?
Thereisnotmuchdifferencebetween10gand11gR(1)RAC.Butthereisasignificantdifferencein11gR2.
Priorto11gR1(10g)RAC,thefollowingweremanagedbyOracleCRS
Databases
Instances
Applications
NodeMonitoring
EventServices
HighAvailability
From11gR2(onwards)itscompletedHAstackmanagingandprovidingthefollowingresourcesasliketheother
clustersoftwarelikeVCSetc.
Databases
Instances
Applications
ClusterManagement
NodeManagement
EventServices
HighAvailability
NetworkManagement(providesDNS/GNS/MDNSDservicesonbehalfofothertraditionalservices)and
SCANSingleAccessClientNamingmethod,HAIP
StorageManagement(withhelpofASMandothernewACFSfilesystem)
Timesynchronization(ratherdependingupontraditionalNTP)
RemovedOSdependenthangcheckeretc,manageswithownadditionalmonitorprocess
Whatishangchecktimer?
[Link]
restartedautomatically.
Thereare2keyparametersforthismodule:
>hangchecktick:[Link]
valueis60secondsOraclerecommendssettingitto30seconds.
>hangcheckmargin:thisdefinesthemaximumhangdelaythatshouldbetoleratedbeforehangchecktimer
resetstheRACnode.
StatetheinitializationparametersthatmusthavesamevalueforeveryinstanceinanOracleRAC
database?
[Link]
[Link]
instancearegivenbelow:
ACTIVE_INSTANCE_COUNT
ARCHIVE_LAG_TARGET
COMPATIBLE
CLUSTER_DATABASE
CLUSTER_DATABASE_INSTANCE
CONTROL_FILES
DB_BLOCK_SIZE
DB_DOMAIN
DB_FILES
DB_NAME
DB_RECOVERY_FILE_DEST
DB_RECOVERY_FILE_DEST_SIZE
DB_UNIQUE_NAME
INSTANCE_TYPE(RDBMSorASM)
PARALLEL_MAX_SERVERS
REMOTE_LOGIN_passWORD_FILE
UNDO_MANAGEMENT
WhatisRAC?WhatisthebenefitofRACoversingleinstancedatabase?
InRealApplicationClustersenvironments,allnodesconcurrentlyexecutetransactionsagainstthesame
[Link]'saccesstotheshareddatatoprovideconsistency
andintegrity.
Benefits:
Improveresponsetime
Improvethroughput
Highavailability
Transparency
AdvantagesofRAC(RealApplicationClusters)
Reliabilityifonenodefails,thedatabasewon'tfail
Availabilitynodescanbeaddedorreplacedwithouthavingtoshutdownthedatabase
Scalabilitymorenodescanbeaddedtotheclusterastheworkloadincreases
WhatisavirtualIPaddressorVIP?
AvirtualIPaddressorVIPisanalternateIPaddressthattheclientconnectionsuseinsteadofthestandard
[Link],weneedtoreserveaspareIPaddressforeachnode,andtheIP
addressesmustusethesamesubnetasthepublicnetwork.
WhatistheuseofVIP?
Ifanodefails,thenthenode'sVIPaddressfailsovertoanothernodeonwhichtheVIPaddresscanacceptTCP
connectionsbutitcannotacceptOracleconnections.
GivesituationsunderwhichVIPaddressfailoverhappens:
VIPaddressesfailoverhappenswhenthenodeonwhichtheVIPaddressrunsfails,allinterfacesfortheVIP
addressfails,allinterfacesfortheVIPaddressaredisconnectedfromthenetwork.
UsingvirtualIPwecansaveourTCP/IPtimeoutproblembecauseOraclenotificationservicemaintains
communicationbetweeneachnodesandlisteners.
WhatisthesignificanceofVIPaddressfailover?
WhenaVIPaddressfailoverhappens,ClientsthatattempttoconnecttotheVIPaddressreceivearapid
[Link]'thavetowaitforTCPconnectiontimeoutmessages.
Whatisvotingdisk?
[Link]
nodesintheclusterregisterstheirheartbeatinformationinthevotingdisk,soastoconfirmthattheyareall
[Link]
[Link](ClusterSynchronizationService)daemonintheclusterwaremaintainstheheart
[Link],thenitwill
rebootitself,thushelpavoidingthesplitbrainsyndrome.
Forhighavailability,Oraclerecommendsthatyouhaveaminimumofthreeoroddnumber(3orgreater)of
votingdisks.
[Link]
ownershipbetweenthenodesincaseoffailure.
TheVotingDiskFilesareusedbyOracleClusterwaretodeterminewhichnodesarecurrentlymembersofthe
[Link]
theclustersintegrity.
[Link]
[Link]
resultCSScancontinuetomaintaintheOracleclustereveniftheASMinstancehasfailed.
Howmanyvotingdisksareyoumaintaining?
[Link]
x
BydefaultOraclewillcreate3votingdiskfilesinASM.
[Link]
configureanoddnumberofvotingdisks>=[Link]
causetheentireclustertofail.
[Link],ifyouareusingASMandexternal
[Link]
redundancyyouwillneed560MB.
Whyweneedtokeepoddnumberofvotingdisks?
[Link]
configureanoddnumberofvotingdisks>=[Link]
causetheentireclustertofail.
WhatareOracleRACsoftwarecomponents?
[Link]
[Link]
GES(GlobalEnqueueService),GCS(GlobalCacheService)[Link]
composedoffollowingbackgroundprocesses:
ACMSAtomicControlfiletoMemoryService(ACMS)
GTX0jGlobalTransactionProcess
LMONGlobalEnqueueServiceMonitor
LMDGlobalEnqueueServiceDaemon
LMSGlobalCacheServiceProcess
LCK0InstanceEnqueueProcess
RMSnOracleRACManagementProcesses(RMSn)
RSMNRemoteSlaveMonitor
WhatareOracleClusterwareprocessesfor10g?
ClusterSynchronizationServices(ocssd)Managesclusternodemembershipandrunsastheoracleuser
failureofthisprocessresultsinclusterrestart.
ClusterReadyServices(crsd)Thecrsprocessmanagesclusterresources(whichcouldbeadatabase,an
instance,aservice,aListener,avirtualIP(VIP)address,anapplicationprocess,andsoon)basedonthe
resource'[Link],stop,monitorandfailover
[Link]
Eventmanagerdaemon(evmd)Abackgroundprocessthatpublisheseventsthatcrscreates.
ProcessMonitorDaemon(OPROCD)ThisprocessmonitortheclusterandprovideI/[Link]
performsitscheck,stopsrunning,andifthewakeupisbeyondtheexpectedtime,thenOPROCDresetsthe
[Link].
OPROCDusesthehangchecktimeronLinuxplatforms.
RACG(racgmain,racgimon)ExtendsclusterwaretosupportOraclespecificrequirementsandcomplex
[Link].
WhatareOracledatabasebackgroundprocessesspecifictoRAC?
LMSGlobalCacheServiceProcess
LMDGlobalEnqueueServiceDaemon
LMONGlobalEnqueueServiceMonitor
LCK0InstanceEnqueueProcess
OracleRACinstancesusetwoprocesses,theGlobalCacheService(GCS)andtheGlobalEnqueueService
(GES).TheGCSandGESmaintainrecordsofthestatusesofeachdatafileandeachcachedblockusinga
GlobalResourceDirectory(GRD).TheGRDcontentsaredistributedacrossalloftheactiveinstances.
WhatisCacheFusion?
[Link]
[Link]
anotherinstanceisinneedofthesameblock,itiseasytogettheblockimagefromtheinsatncewhichhasthe
[Link]
[Link](GES)monitorsandInstanceenqueueprocessmanagesthe
cahcefusion
WhatisSCAN?(11gR2feature)
SingleClientAccessName(SCAN)issanewOracleRealApplicationClusters(RAC)11gRelease2feature
[Link]
usingSCANdonotneedtochangeifyouaddorremovenodesinthecluster.
SCANprovidesasingledomainnamevia(DNS),allowinganduserstoaddressaRACclusterasifitwerea
[Link](VIP).
Singleclientaccessname(SCAN)ismeanttofacilitatesinglenameforallOracleclientstoconnecttothe
clusterdatabase,[Link],wehavetokeepadding
[Link],whenanewnodegetsaddedtoordeletedfromthe
cluster.
SingleClientAccessName(SCAN)eliminatestheneedtochangeTNSNAMESentrywhennodesareaddedto
[Link]
recommendsassigning3addressestoSCAN,whichwillcreate3SCANlisteners,thoughtheclusterhasgot
dozensofnodes..SCANisadomainnameregisteredtoatleastoneanduptothreeIPaddresses,eitherin
DNS(DomainNameService)orGNS(GridNamingService).TheSCANmustresolvetoatleastoneaddress
[Link],OraclerecommendsconfiguringtheSCANtoresolve
tothreeaddresses.
[Link]
WhatareSCANcomponentsinacluster?
[Link]
[Link](3)
[Link](3)
WhatisFAN?
FastapplicationNotificationasitabbreviatestoFANrelatestotheeventsrelatedtoinstances,servicesand
[Link]
andservicelevelinformationthatincludesservicestatuschangessuchas,[Link]
canrespondtoFANeventsandtakeimmediateaction.
WhatisTAF?
TAF(TransparentApplicationFailover)isaconfigurationthatallowssessionfailoverbetweendifferentnodes
ofaRACdatabasecluster.
TransparentApplicationFailover(TAF).Ifacommunicationlinkfailureoccursafteraconnectionisestablished,
[Link],andsession
[Link],ifthestatementexecutingatthetimeof
thefailoverisaSelectstatement,thatstatementmaybeautomaticallyreexecutedonthenewconnectionwith
thecursorpositionedontherowonwhichitwaspositionedpriortothefailover.
AfteranOracleRACnodecrashesusuallyfromahardwarefailureallnewapplicationtransactionsare
[Link]
were"inflight"[Link]
torestartinflightapplicationtransactions,allowingafailednodetoresumeprocessingonanotherserver
[Link]'sanswertoapplicationfailoverisanewOracleNetmechanismdubbed
[Link]
OracleNetclient.
TAFarchitectureofferstheabilitytorestarttransactionsateitherthetransaction(SELECT)orsessionlevel.
WhataretherequirementsforOracleClusterware?
[Link](VotingDiskandOracleClusterRegistryOCR)
[Link](andthreesetofIPaddress)
NetworkCard1(withIPaddressset1)forpublicnetwork
NetworkCard2(withIPaddressset2)forprivatenetwork(forinternodecommunicationbetweenracnodes
usedbyclusterwareandracdatabase)
IPaddressset3forVirtualIP(VIP)(usedasVirtualIPaddressforclientconnectionandforconnectionfailover)
[Link],OCFS2(OracleClusterFileSystem),NFS,..
WhichenabletheloadbalancingofapplicationsinRAC?
OracleNetServicesenabletheloadbalancingofapplicationconnectionsacrossalloftheinstancesinan
OracleRACdatabase.
HowtofindlocationofOCRfilewhenCRSisdown?
IfyouneedtofindthelocationofOCR(OracleClusterRegistry)butyourCRSisdown.
WhentheCRSisdown:
[Link],locationofthisfilechangesdependingontheOS:
OnLinux:/etc/oracle/[Link]
OnSolaris:/var/opt/oracle/[Link]
WhenCRSisUP:
SetASMenvironmentorCRSenvironmentthenrunthebelowcommand:
ocrcheck
In2nodeRAC,howmanyNICsarerusing?
2networkcardsoneachclusterwarenode
NetworkCard1(withIPaddressset1)forpublicnetwork
NetworkCard2(withIPaddressset2)forprivatenetwork(forinternodecommunicationbetweenracnodes
usedbyclusterwareandracdatabase)
In2nodeRAC,howmanyIPsarerusing?
63setofIPaddress
##eth1Public:2
##eth0Private:2
##VIP:2
HowtofindIPsinformationinRAC?
Editthe/etc/hostsfileasshownbelow:
#Donotremovethefollowingline,orvariousprograms
#thatrequiresnetworkfunctionalitywillfail.
[Link]
##PublicNodenames
[Link].netnode1pub
[Link].netnode2pub
##PrivateNetwork(Interconnect)
192.168.0.11node1prvnode1prv
192.168.0.22node2prvnode2prv
##PrivateNetwork(NetworkAreastorage)
192.168.1.11node1nasnode1nas
192.168.1.22node2nasnode2nas
192.168.1.33nasservernasserver
##VirtualIPs
[Link].netnode1vip
[Link].netnode2vip
WhatisdifferencebetweenRACipaddresses?
PublicIPadressisthenormalIPaddresstypicallyusedbyDBAandSAtomanagestorage,systemand
[Link].
PrivateIPaddressisusedonlyforinternalclusteringprocessing(CacheFusion)(akaasinterconnect).Private
IPaddressesarereservedforprivatenetworks.
[Link]
VIPissoclientconnectioncanbefailovertosurvivingnodesincasethereisfailure
Canapplicationdeveloperaccesstheprivateip?
[Link](CacheFusion)(akaasinterconnect)
[Link]
WhatisRAC?
RACArchitechture?
WhatisaSCANListener?
WhatisVIP?
WhatisTAF?
WhatisFAN?
WhatisLMON?
WhatisCACHEFUSION?
WhatisSPLITBRAINinRAC?
HowLoadbalancingworksinORACLERAC?
WhatisVotingDisk?
HowmanybackupsarethereforVotingDisk?
WHYODDNUMBEROFVOTINGDISKISUSED?
HowtotakeVotingDiskbackup?DoyoutakeVotingDiskbackup?
IfyoulostVotingDisk,howdoyouRestore?
WhatisOCR?
HowmanybackupsarethereforOCR?
HowtotakeOCRbackup?DoyoutakeOCRbackup?
IfyoulostOCR,howdoyouRestore?
WhatisaMasterNode?
HowdoyouknowwhichistheMasterNodewithoutconnectingtoDatabase?
CanwehaveNode1inSOLARISOSandNode2inLINUXOS?
CommomnlyusedcommandsinRAC?
HowtoconvertstandalonedatabasetoRAC?