1. Get rid of all advertisements and get unlimited access to documents by upgrading to Premium Membership. Upgrade to Premium Now and also get a Premium Badge!

Cloning a Data Guard ASM instance

Discussion in 'Security, Backup and Recovery' started by General Franco, Aug 10, 2012.

  1. General Franco

    General Franco Active Member

    Messages:
    22
    Likes Received:
    1
    Trophy Points:
    90
    I'm cloning an Oracle primary dataguard instance out of ASM and onto UFS. When I move it onto UFS I want to disable all standby activity and switch it to run as a regular DB.
    Before I restore the control files from the backup and move them to UFS I switch all the spfile params that ref dataguard e.g.

    alter system set dg_broker_start=FALSE scope=spfile;
    alter system set log_archive_config='' scope=spfile;
    alter system set log_archive_dest_1='LOCATION=/data/new_location
    alter system set log_archive_dest_2='' scope=spfile;
    alter system set log_archive_dest_state_1 = 'defer';
    alter system set log_archive_dest_state_2 = 'defer';
    alter system set log_archive_dest_state_3 = 'defer';
    alter system set log_archive_dest_3='' scope=spfile;
    alter system set standby_archive_dest='' scope=spfile;
    alter system set standby_file_management=MANUAL scope=spfile;

    Then I restore the datafiles and arch logs to the new UFS locations.
    Switch datafile all;
    Rename the online redo.
    Afterwards I try to recover the DB to the latest scn.

    With regular DBs this works fine .. however with the RMAN backup set I get from the primary DG instance I get the following failure whenI try to 'alter database open resetlogs':

    RMAN-03002: failure of alter db command at 08/10/2012 10:02:39
    ORA-01092: ORACLE instance terminated. Disconnection forced
    ORA-00704: bootstrap process failure
    ORA-39700: database must be opened with UPGRADE option

    The RDBMS & RHEL versions are the same on both source and target machines. Any ideas?
    I went ahead and upgraded the target DB using catupgrd.sql and it runs fine .. but that's not how it should work .. right?


    Many Thanks.
     
  2. zargon

    zargon Community Moderator Forum Guru

    Messages:
    2,345
    Likes Received:
    347
    Trophy Points:
    1,430
    Location:
    Aurora, CO
    Depends. Which release of Oracle?
     
  3. General Franco

    General Franco Active Member

    Messages:
    22
    Likes Received:
    1
    Trophy Points:
    90
    The release is 11.2.0.3
     
  4. zargon

    zargon Community Moderator Forum Guru

    Messages:
    2,345
    Likes Received:
    347
    Trophy Points:
    1,430
    Location:
    Aurora, CO
    Does your data guard installation use the broker? If so this can happen as Oracle believes the new standalone to be of a different version. There is no harm in what you did and it makes Oracle happy. Had you not used the broker (and I am presuming you are even though you haven't yet answered my question) this probably wouldn't be an issue and the db would startup without error.
     
  5. General Franco

    General Franco Active Member

    Messages:
    22
    Likes Received:
    1
    Trophy Points:
    90
    Thanks for the feedback. Yes the source DB was using the Broker. So is this a bug?
    Is there anyway of avoiding having to run an upgrade?

    Also .. I didn't have to issue the 'ALTER DATABASE RECOVER MANAGED STANDBY DATABASE CANCEL;' cmd .. which is normal practice in these circumstances (according to the Oracle UFS/ASM clone documentation). Seems like odd behaviour.