2 Replies Latest reply: Aug 20, 2009 3:50 AM by 539769 RSS

    online logs needed archiving

    609621
      we have 10.2.0.4 on solaris 5.10

      Instance hanged coz of this reason:


      MED1 - Can not allocate log, archival required
      Tue Aug 18 09:46:05 2009
      Thread 1 cannot allocate new log, sequence 53164
      All online logs needed archiving


      so,we made the below change in the environment :





      increased log_archive_max_process=10
      increased redlogs form 300Mb to 500Mb with additional groups..


      Please advise if this is the correct solution ?


      Kai
        • 1. Re: online logs needed archiving
          asifkabirdba
          Is your DB in archive log mode? If enabled do you have enough space for archival destination? If you archival destination is full then delete few archivelogs to reclaim some space.

          OR

          If your redo log size is too small to handle your current redo log switch frequency then bigger redo log size will work in this case. Increase redo log size to 600M.


          OTHERWISE

          If automatic archival is disabled, run the following command

          ALTER SYSTEM ARCHIVE LOG ALL;


          it is better if you set as well LOG_ARCHIVE_START parameter into init.ora

          LOG_ARCHIVE_START=TRUE


          Regards
          Asif Kabir
          • 2. Re: online logs needed archiving
            539769
            MED1 - Can not allocate log, archival required
            It seems that all the online redo logs are full and you need to empty them. This can happeen due to two reasons
            1) archivelog destination is full
            2)archiving is disabled although the database is in archive log mode
            increased log_archive_max_process=10
            increased redlogs form 300Mb to 500Mb with additional groups..
            Please advise if this is the correct solution ?
            These solution will not help you. See if the sufficient empty space is there at log_archive_destination. If it is there it may happen that archive log files were removed by OS commands and database has no knowledge about the removal. In this case connect to RMAN and crosscheck archivelogs, delete the expired. This will let the database archive all redo logs once you execute the command 'ALTER SYSTEM SWITCH LOGFILE' . If enough space is not there backup the old archivelog and delete them as described above. If archiving is disabled the enable them executing the command 'ALTER SYSTEM set log_archive_dest_state_1=enable'

            Rgds.