Re: Crash recovery after shutdown immediate?

On Dec 9, 9:43 am, Helma <helma.vi..._at_hotmail.com> wrote:
> On Dec 9, 2:32 pm, hpuxrac <johnbhur..._at_sbcglobal.net> wrote:
>
>
>
>
>
> > On Dec 9, 4:05 am, Helma <helma.vi..._at_hotmail.com> wrote:
>
> > > Hello *,
>
> > > Odd problem i had after i had shut down the database (immediate) and
> > > made a media copy. After restoring and opening with resetlogs, the
> > > database stated it wanted a recovery. After i did a recover until
> > > cancel, and cancelled immediately on first occasion, the database
> > > started.
>
> > > I ignored this incident, but now i am curious about this entry in the
> > > logfile. Yesterday, i did again a shutdown immediate ( i never do
> > > abort) ,  made a copy of the database and restarted the database. In
> > > the errorlog, it seems that a crash recovery occured. This shoudn't
> > > happen? What am i missing?
>
> > > ALTER DATABASE OPEN
> > > Mon Dec  8 18:03:40 2008
> > > Beginning crash recovery of 1 threads
> > > Mon Dec  8 18:03:40 2008
> > > Started redo scan
> > > Mon Dec  8 18:03:40 2008
> > > Completed redo scan
> > >  156 redo blocks read, 66 data blocks need recovery
> > > Mon Dec  8 18:03:40 2008
> > > Started recovery at
> > >  Thread 1: logseq 58, block 16913, scn 0.0
> > > Mon Dec  8 18:03:40 2008
> > > Recovery of Online Redo Log: Thread 1 Group 3 Seq 58 Reading mem 0
> > >   Mem# 0 errs 0: /oracle/log_a/xxx/log3A.log
> > >   Mem# 1 errs 0: /oracle/log_b/xxx/log3B.log
> > > Mon Dec  8 18:03:40 2008
> > > Completed redo application
> > > Mon Dec  8 18:03:41 2008
> > > Ended recovery at
> > >  Thread 1: logseq 58, block 17069, scn 0.4242238601
> > >  66 data blocks read, 66 data blocks written, 156 redo blocks read
> > > Crash recovery completed successfully
>
> > Where's the log output from the shutdown immediate that you are
> > claiming completed successfully?
>
> > My shot in the dark is that you either did not get a good shutdown
> > ( but asssumed that you did ) ... or that you restored something
> > ( else ) that was not from a clean shutdown.- Hide quoted text -
>
> > - Show quoted text -
>
> Hello hpuxrac,
>
> the log output from the shutdown immediate:
>
> ARC0: Completed archiving  log 2 thread 1 sequence 57
> Mon Dec  8 15:31:10 2008
> Shutting down instance: further logons disabled
> Shutting down instance (immediate)
> License high water mark = 39
> Mon Dec  8 15:33:01 2008
> ALTER DATABASE CLOSE NORMAL
> Mon Dec  8 15:33:02 2008
> SMON: disabling tx recovery
> SMON: disabling cache recovery
> Mon Dec  8 15:33:18 2008
> Shutting down archive processes
> Archiving is disabled
> Mon Dec  8 15:33:18 2008
> ARCH shutting down
> Mon Dec  8 15:33:18 2008
> ARC0: Archival stopped
> Mon Dec  8 15:33:18 2008
> ARCH shutting down
> Mon Dec  8 15:33:18 2008
> ARC1: Archival stopped
> Mon Dec  8 15:33:24 2008
> Thread 1 closed at log sequence 58
> Successful close of redo thread 1
> Mon Dec  8 15:33:49 2008
> Completed: ALTER DATABASE CLOSE NORMAL
> Mon Dec  8 15:33:49 2008
> ALTER DATABASE DISMOUNT
> Completed: ALTER DATABASE DISMOUNT
> ARCH: Archiving is disabled
> Shutting down archive processes
> Archiving is disabled
> Archive process shutdown avoided: 0 active
> ARCH: Archiving is disabled
> Shutting down archive processes
> Archiving is disabled
> Archive process shutdown avoided: 0 active
>
> After that, I did a backup, NOT a restore. I restarted the database
> and then the crash recovery appeared - very odd. Oracle 9.2.0.8  64-
> bits AIX
>
> I will soon have to restore, i'm curious what will happen...- Hide quoted text -
>
> - Show quoted text -

Your initial post said you did a "restore" not a backup. Why did you open reset logs intead of just opening the database normally? Reset logs is usually on issued after a restore.

Do a shutdown immedate, make a cold backup, and restart normally and you will not see any errors or crash recovery messages since the control file, database files, and online redo log would all be in sync.

HTH -- Mark D Powell -- Received on Tue Dec 09 2008 - 09:48:43 CST