DBAsupport.com Forums - Powered by vBulletin
Results 1 to 2 of 2

Thread: Data Guard Switchover Problem

Hybrid View

  1. #1
    Join Date
    Jun 2007
    Posts
    3

    Data Guard Switchover Problem

    During a recent test of our switchover process on our production environment, the SQL in the log files was not applied to the new logical
    standby database. The archives logs are transferred to the logical standby, but the SQL is not applied. We had to perform a manual process to keep the
    databases in sync. We also had problems with the switchback because of this problem. The manual workaround process is described below:

    Run the following query--
    spool status.txt
    set pagesize 1000
    SELECT L.thread#, L.sequence#, L.first_time,
    (CASE WHEN L.next_change# WHEN L.first_change# < p.applied_scn THEN 'CURRENT'
    ELSE 'NO' END) APPLIED
    FROM DBA_LOGSTDBY_LOG L, DBA_LOGSTDBY_PROGRESS P
    WHERE L.first_time > sysdate - 3
    ORDER BY thread#, SEQUENCE#

    spool off
    ;

    THREAD# SEQUENCE# FIRST_TIME APPLIED
    ---------- ---------- -------------------- -------
    1 62481 Jun-02-2007 18:20:22 NO
    1 62482 Jun-02-2007 18:30:05 NO
    1 62483 Jun-02-2007 18:30:09 NO
    1 62484 Jun-02-2007 18:31:02 NO
    1 62485 Jun-02-2007 18:51:04 NO
    1 62486 Jun-02-2007 19:11:06 NO
    2 61123 Jun-02-2007 18:20:15 NO
    2 61124 Jun-02-2007 18:30:05 NO
    2 61125 Jun-02-2007 18:30:09 NO
    2 61126 Jun-02-2007 18:32:27 NO
    2 61127 Jun-02-2007 18:33:27 NO
    2 61128 Jun-02-2007 18:53:27 NO
    2 61129 Jun-02-2007 19:13:28 NO

    13 rows selected.

    2. Delete archive log with the latest sequence number from cluster1. The example this would be --62486. The log is located in /u03/oracle/archive/
    and is the latest log from the Primary database cluster 1.
    3. The next step is to register the logfile using the following command alter database register
    logical logfile '/u03/oracle/archive/standby/log_instance_1_
    62486.arc'
    4. The process changes the no's to current or yes. The process has to be repeated as
    the logs are transferred to the new standby to keep the databases in sync.

    *********************************************************

    Some of the errors that we saw during the switchover are listed below:



    The following error started after completion of the alter database commit to switchover to
    logical primary command completed successfully:


    Error Group 1
    ORA-02068: following severe error from instance2
    ORA-03113: end-of-file on communication channel

    The Instance1 database was generating an error talking to Instance2. There is a database link
    configured for communication between Instance1 and and Instance2 (RAC)
    Both databases have a matching standby database. Instance1_STB and Instance2_STB (non-RAC)

    Errors Group 2 ( on instance 1 & 2)

    ORA-00321: log 1 of thread 1, cannot update log file header
    ORA-00312: online log 1 thread 1: '/u01/oracle/oradata/instance2/log1_01a.rdo'
    ORA-17500: ODM err:ODM ERROR V-41-4-1-83-9 Bad file number

    Has anyone had this type of problem where after you switchover you cannot get the new standby to automatically apply logs.

  2. #2
    Join Date
    Jun 2006
    Posts
    259
    I'm a bit confused:
    Is Instance1, Instance2 part of a cluster? (ie RAC)

    And Instance1_STB, Instance2 part of a non cluster?

    If this is the case then why have instance2 ?

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  


Click Here to Expand Forum to Full Width