This discussion is archived
5 Replies Latest reply: Jan 25, 2013 12:31 AM by Sebastian Solbach (DBA Community) RSS

RAC Service not failing over when Preferred Node goes down

Tom Newbie
Currently Being Moderated
GI version : 11.2.0.3
RDBMS version : 11.2.0.3

Platform : RHEL 5.4


We create service using the command shown below and then start it.
$ srvctl add service -d SDLJMR -s CPS_TST_02 -r SDLJMR1 -a SDLJMR2
The clients are able to connect using the service. But when the preferred node goes down for any reason the service is not failed over to the available node. If this is the case why configure "Available" node ?

Is there anything else we have to configure to get the service to failover to the surviving node ?
  • 1. Re: RAC Service not failing over when Preferred Node goes down
    JohnWatson Guru
    Currently Being Moderated
    Can you confirm the situation?
    srvctl config service -d SDLJMR -s CPS_TST_02 -v
    srvctl status service -d SDLJMR -s CPS_TST_02 -v
  • 2. Re: RAC Service not failing over when Preferred Node goes down
    Tom Newbie
    Currently Being Moderated
    Output of the query you've asked for
    $ srvctl config service -d SDLJMR -s CPS_TST_02 -v
    Service name: CPS_TST_02
    Service is enabled
    Server pool: SDLJMR_CPS_TST_02
    Cardinality: 1
    Disconnect: false
    Service role: PRIMARY
    Management policy: AUTOMATIC
    DTP transaction: false
    AQ HA notifications: false
    Failover type: NONE
    Failover method: NONE
    TAF failover retries: 0
    TAF failover delay: 0
    Connection Load Balancing Goal: LONG
    Runtime Load Balancing Goal: NONE
    TAF policy specification: NONE
    Edition:
    Preferred instances: SDLJMR1
    Available instances: SDLJMR2
    
    
    
    $ srvctl status service -d SDLJMR -s CPS_TST_02 -v
    Service CPS_TST_02 is running on instance(s) SDLJMR1
    Last week we were doing a rolling patch and we had to bring down Node1. But the service didn't failover to Node2 where the available instance is
  • 3. Re: RAC Service not failing over when Preferred Node goes down
    rcc50886 Journeyer
    Currently Being Moderated
    Last week we were doing a rolling patch and we had to bring down Node1. But the service didn't failover to Node2 where the available instance is
    The service will not move to available instance if we Manually shutdown the instance. The service will start on available node only if there is any unplanned outage like node ejection...

    And also remember, we manually relocate the service to original node once it back online.

    -Thanks
  • 4. Re: RAC Service not failing over when Preferred Node goes down
    Tom Newbie
    Currently Being Moderated
    The service will not move to available instance if we Manually shutdown the instance.

    Oh Ok. I hope in future versions ( >=12c ) , they implement it.
  • 5. Re: RAC Service not failing over when Preferred Node goes down
    Sebastian Solbach (DBA Community) Guru
    Currently Being Moderated
    Hi Tom,

    I doubt it, this is works as desgned. (I believe they introduced this behaviour with 11.2.0.2).
    Main reason is: If you say you want to shutdown the instance, then you also want to shutdown the services running in that instance. If you don't want that, then manually relocate the service, before shutdown.

    In any case if the datbase crashes it will definitely failover.

    Regards
    Sebastian

Legend

  • Correct Answers - 10 points
  • Helpful Answers - 5 points