This discussion is archived
1 2 Previous Next 19 Replies Latest reply: Jan 23, 2013 6:38 AM by 802461 Go to original post RSS
  • 15. Re: Multiple instance at same server with same ORACLE_HOME and Same SID
    802461 Newbie
    Currently Being Moderated
    Hello, this is the alert for standby


    Completed: alter database add standby logfile '/disk/prod/DR/2/oradata/prodrot/srl10' size 4m    --------> here is the last file before recover command , ok until here
    Fri Dec 21 15:15:59 2012
    ALTER DATABASE RECOVER  managed standby database disconnect ------------------> recover command wich broke the production database, nothing broke on standby
    Attempt to start background Managed Standby Recovery process (prodrot)
    Fri Dec 21 15:15:59 2012
    MRP0 started with pid=22, OS id=10773
    MRP0: Background Managed Standby Recovery process started (prodrot)
    started logmerger process
    Fri Dec 21 15:16:04 2012
    Managed Standby Recovery not using Real Time Apply
    Parallel Media Recovery started with 4 slaves
    Fri Dec 21 15:16:05 2012
    Warning: recovery process PR01 cannot use async I/O
    Fri Dec 21 15:16:05 2012
    Warning: recovery process PR02 cannot use async I/O
    Fri Dec 21 15:16:05 2012
    Warning: recovery process PR03 cannot use async I/O
    Fri Dec 21 15:16:05 2012
    Warning: recovery process PR04 cannot use async I/O
    Waiting for all non-current ORLs to be archived...
    All non-current ORLs have been archived.
    Errors in file /oracle/app/oracle_prod/rdbms/log/diag/rdbms/prodrot/prodrot/trace/prodrot_mrp0_10773.trc:
    ORA-00367: checksum error in log file header
    ORA-00305: log 11 of thread 1 inconsistent; belongs to another database
    ORA-00312: online log 11 thread 1: '/disk/prod/2/oradata/prod/rlog11' ----------> this is what i´m saying, standby, with different path tried to use rlog on prod path.
    Clearing online redo logfile 11 /disk/prod/2/oradata/prod/rlog11
    Clearing online log 11 of thread 1 sequence number 992640 -------> as this happen and standby start using it, prod crashs.
    Errors in file /oracle/app/oracle_prod/rdbms/log/diag/rdbms/prodrot/prodrot/trace/prodrot_mrp0_10773.trc:
    ORA-00367: checksum error in log file header
    ORA-00305: log 11 of thread 1 inconsistent; belongs to another database ------------> again and this repeat for all rlogs in prod.
    ORA-00312: online log 11 thread 1: '/disk/prod/2/oradata/prod/rlog11'
    Fri Dec 21 15:16:52 2012
    Media Recovery Waiting for thread 1 sequence 992292 -----------> standby start working normally but master crashed at this time.
  • 16. Re: Multiple instance at same server with same ORACLE_HOME and Same SID
    mseberg Guru
    Currently Being Moderated
    What is appears to be is internal Oracle thinks the log file exist where the primary log files exist and thus the conflict.

    Since a Standby database does not use Redo logs and Standby are optional you should in theory be able to drop and recreate them and stop the conflict.

    I have never done anything like this and this note is not the glass full I'm looking for.


    ORA-305 ora-312 trying to open the database [ID 1326538.1]


    The completely safe option is to recreate the standby.

    I hate to try something I have not tested.

    Best Regards

    mseberg
  • 17. Re: Multiple instance at same server with same ORACLE_HOME and Same SID
    802461 Newbie
    Currently Being Moderated
    thanks again for reply.

    yes what i think is that as controlfile is a standby creation from production control it try to at least lock all the files needed in case to switch over to primary could be ???? i forget to tell this is a physical standby but you guys know already.

    so the problem just happen because the master was in the same node, if disks are not avaiable, standby ignore the error and go ahead and start applying the logs that´s what happens before on 2 instances i made standby at this server, the problem rises when master was in this server so disks avaiable, and i was implementing another physical standby there, so when i put in recover it distroy the production :(.

    about standby i´m not worried, i can rebuild it anytime my question is how to proceed, should i before put in recover, rename the rlogs like :

    alter database rename file '/disk/prod/2/oradata/prod/rlog11' to '/disk/prod/DR/2/oradata/prodtst/rlog11';

    i´ll not bring physical standby when prod is running at this node anymore, but what will happen if we try to open the database if it still thinks that rlogs are on wrong path? i think standby will not be in good shape to use in case needed.

    thanks to reply.

    br
  • 18. Re: Multiple instance at same server with same ORACLE_HOME and Same SID
    802461 Newbie
    Currently Being Moderated
    Hello all, i solved the problem by setting the *.log_file_name_convert to the correct place. i missunderstand documentation and i was thinking that was for archived redo logs and that is for online redo log wich is the same path of my datafiles.
    also another way to solve is the way i already told before by manually renaming the files.

    now is all set with new standby´s

    thanks.
1 2 Previous Next

Legend

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