11gR2 RAC Database Server Maintenace

 Hi there,

      Here I have presented the maintenance steps of 11gR2 RAC. Often we wonder about the certain steps to follow if we require any kind of upgradation that has to be done on the server running RAC Databases.

If you will not follow certified steps for maintenance then you may have to pay for this. So here I have tried to present a completely logical way to do accomplish the activities.    

Note: Before Maintenance All the  Resources( A resource could be an application process, a database, a service, a listener, and so on.) should be down.

CLUSTER DB SERVER MAINTENANCE IN 11GR2 RAC

============================== =============

ACTIVITY PLAN: FIRMWARE UPGRADATION,OS PATCHING,H/W UPGRADATION AND OTHER MAINTENANCE

DOWNTIME TO DATABASE: 0 DOWNTIME

NOTE: MAINTENANCE ACTIVITIES WILL BE DONE ONE NODE FOLLOWED BY ANOTHER

CLUSTER DATABASE SHUTDOWN PROCESS:

Let’s say we need to do maintenance activity on the server one by one in order to avoid downtime.

Here in our case ehishchn1 is the Node1 instance

STEP-1: DISABLE THE INSTANCE (This Step will update the flag to disable So that next time after server restart instance will not start automatically)

    $srvctl disable instance -d ehischn -i ehischn1

STEP-2: STOP INSTANCE IN NODE1

    $srvctl stop instance -d ehischn -i ehischn1

STEP-3: DISABLE CLUSTER IN NODE1

   Go to the $GRID_HOME/bin

   #./crsctl disable crs

STEP-4: STOP CRS IN NODE1 Then Check for listener and also pmon services to reconfirm

  #./crsctl stop crs

  #ps -ef|grep tns

  #ps -ef|grep pmon

STEP-5: RESTART THE NODE1 (THIS IS REQUIRED BECAUSE THERE COULD BE SOME CLUSTER SERVICES WHICH HAVE NOT GONE DOWN)

  #reboot

Note: After reboot recheck for pmon service using ps -ef|grep pmon 

 

STEP-6: DO THE MAINTENANCE THEN RESTART THE SERVER

STEP-7: ENABLE THE CLUSTER NODE1

 #./crsctl enable crs

STEP-8:ENABLE THE INSTANCE

 $srvctl enable database -d ehischn -i ehischn1

STEP-9: START THE CLUSTER

 #./crsctl start crs (This will start the cluster as well instance up if it will not startup the instance then use srvctl start instance command to start the instance )

############### FOR STORAGE MAINTENACE########

NEED TO DOWN ALL THE SERVERS

NEED TO TAKE BACKUP TO SEPARATE STORAGE 

DOWN DATBASE

$srvctl stop database -d ehischn (This will stop database all instances in clusterwide)

DOWN CLUSTERWIDE SERVICES(CLUSTER SERVICES RUNNING ON ALL NODES)

#./crsctl stop cluster -all (It will stop all cluster services)

AFTER STORAGE MAINTENANCE ACTIVITIES START THE SERVERS, ALL NODES CRS AS WELL AS DB SERVICES WILL COME UP AUTOMATICALLY AS WE HAVE NOT DISABLED ANY SERVICES EXPLICITLY

Hope this document will help you guys.

Please comment and suggest if any improvement to this blog required.