Dear Boss Riky, Metalink *Note:438176.1*
Applies to: Oracle Server - Enterprise Edition - Version: 10.1.0.4 This problem can occur on any platform. Symptoms The following error can occur when Oracle is doing any type of recovery that requires the redo logs. The type of recovery can be internal block recovery, database recovery, or standby database recovery. Redo log file corruption reported in the database. ORA-00355: change numbers out of order ORA-00353: log corruption near block 2879362 change 5677245003168 time 06/15/2007 17:09:30 ORA-00312: online log 2 thread 1: '/apps/oracle/EDR1P/u07/data/redo_g2_t1_02.log' ORA-00312: online log 2 thread 1: '/apps/oracle/EDR1P/u06/data/redo_g2_t1_01.log' Changes The error occurred when an import was canceled and Oracle in the same time frame had to do a internal block recovery. Cause This error occurs when a rollback block address (RBA) checkpoint is in the log after the incremental checkpoint that the block should be is has already been done. What are incremental checkpoints? Incremental checkpoints Avoid large peaks of I/O activity at full checkpoints Write the oldest blocks only, instead of all blocks Are only recorded in the control file, avoiding data file header updates >From Oracle8 onwards, this was replaced by the checkpoint queue. It orders the blocks with "oldest first." The checkpoint queue solves two problems. Firstly it enables the incremental checkpoint to avoid having a cache filled with many dirty blocks, which all must be written at once when a checkpoint occurs. Secondly, by keeping the dirty block count low, the number of blocks that need to be recovered in case of a crash is fewer, resulting in faster database recovery. The length of this list is an indicator of the amount of blocks that need recovery if the block buffer is lost in an instance crash. . When the list gets "too long," the DBWn process starts writing blocks. (The block may remain in the buffer as a read copy). The process utilizing incremental checkpoint has changed significantly between Oracle8, Oracle8i, and Oracle9i. In Oracle8 incremental checkpoint was regulated by the DB_BLOCK_MAX_DIRTY_TARGET, and simply consisted of the number of dirty blocks in the cache is kept below a threshold. In Oracle8i incremental checkpoint was regulated by the FAST_START_IO_TARGET, which limited the number of blocks waiting to be needed for recovery. Solution Similar bugs: Unpublished BUG: 3024358 10.1.0.1 RDBMS 10.1.0.1 DATAGUARD_PSBY PRODID-5 PORTID-46 ORA-355 Abstract: ONLINE REDO LOG CORRUPTION DUE TO CHANGE NUMBER OUT OF ORDER Special Parameters: Parameter: DISABLE_INCREMENTAL_CHECKPOINTS ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Disable incremental checkpoints for thread recovery. The incremental checkpoint may be advanced even though there are still buffers in the cache with a low RBA before the checkpoint. This can lead to unexpected errors during recovery. _disable_incremental_checkpoints = true ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Add this parameter to the spfile before recovery of the database. This parameter can be removed after the recovery is completed and the database it open. If needed, the parameter can be left since it is only for incremental checkpoints and not checkpoints at log switches. Thanks & Regards, Henri Simatupang On Fri, May 2, 2008 at 4:22 PM, Riky Kurniawan <[EMAIL PROTECTED]> wrote: > Dear Oraclers, > > ketika saya startup database > > Total System Global Area 536870912 bytes > Fixed Size 789756 bytes > Variable Size 321909508 bytes > Database Buffers 213909504 bytes > Redo Buffers 262144 bytes > Database mounted. > ORA-01113: file 1 needs media recovery > ORA-01110: data file 1: > 'D:\ORACLE\PRODUCT\10.1.0\ORADATA\ORCL\SYSTEM01.DBF' > > lalu saya coba jalankan: > SQL> recover database; > yang muncul malah > > ORA-00283: recovery session canceled due to errors > ORA-00399: corrupt change description in redo log > ORA-00353: log corruption near block 3 change 33592856636 time 05/02/2008 > 15:37:30 > ORA-00312: online log 3 thread 1: > 'D:\ORACLE\PRODUCT\10.1.0\ORADATA\ORCL\REDO03.LOG' > > Adakah rekan² yang pernah mengalami masalah ini? bagaimana solusinya agar > database bisa open secara normal? > > Ada referensi URL yang membantu? > > terima kasih sebelum dan sesudahnya > > [Non-text portions of this message have been removed] > > > [Non-text portions of this message have been removed] ------------------------------------ -- -----------I.N.D.O - O.R.A.C.L.E--------------- Keluar: [EMAIL PROTECTED] Website: http://indooracle.wordpress.com ----------------------------------------------- Bergabung dengan Indonesia Thin Client User Groups, Terminal Server, Citrix, New Moon Caneveral, di: http://indo-thin.blogspot.comYahoo! Groups Links <*> To visit your group on the web, go to: http://groups.yahoo.com/group/indo-oracle/ <*> Your email settings: Individual Email | Traditional <*> To change settings online go to: http://groups.yahoo.com/group/indo-oracle/join (Yahoo! ID required) <*> To change settings via email: mailto:[EMAIL PROTECTED] mailto:[EMAIL PROTECTED] <*> To unsubscribe from this group, send an email to: [EMAIL PROTECTED] <*> Your use of Yahoo! Groups is subject to: http://docs.yahoo.com/info/terms/
