Re: Killing pmon service - does Oracle backout?

Sometimes you can't even open sqlplus or svrmgrl to issue a shutdown abort. He may have been getting a ora-3113 error. In this case you would have to kill the oracle processes to bring down the instance or remove the shared memory segments and semaphores.

>From: Stephane Faroult <sfaroult_at_oriole.com>
>Reply-To: ORACLE-L_at_fatcity.com
>To: Multiple recipients of list ORACLE-L <ORACLE-L_at_fatcity.com>
>Subject: Re: Killing pmon service - does Oracle backout?
>Date: Tue, 16 Jan 2001 11:25:21 -0800
>
>"Barker, Tony" wrote:
>
> > Hi!!,
> >
> > I recently hosed our test system and could not get into any
>utility
> > applications (SQLPLUS, SVRMGRL, ..) and I needed to bring down some hung
> > instances. (The system is Sun Solaris release 2.6 using Oracle 8.1.6) I
> > ended up canceling the PMON task of the various instances to get them to
> > come down. They all terminated and I was sure that I was going to have
>to
> > do a recovery to get them to come back up. To my surprise all the
>instances
> > came up fine after getting everything functionable.
> >
> > Does anyone know what happens when you cancel PMON that allowed
>this
> > to happen? Did SMON clean things up before the instances terminated? I
> > hope I don't have a false sense of comfort!?
> >
> > Thanks in advance for your input!!!
> >
> > Tony Barker
> > DBA - State of Indiana
> > tbarker_at_doit.state.in.us (317)232-0719
> > --
> > Please see the official ORACLE-L FAQ: http://www.orafaq.com
> > --
> > Author: Barker, Tony
> > INET: tbarker_at_doit.state.in.us
> >
>
>Tony,
>
> I would have found 'shutdown abort' cleaner than killing PMON.
>You didn't have any recovery to do because it's automatic ... Very briefly
>the
>redo log files contain the delta between memory, where everything happens,
>and
>datafiles, the 'persistent' part of your DB - at least as far as committed
>transactions are concerned. You need to do a recovery when the delta is
>MORE
>than the current redo log files, for instance because you had to restore an
>old
>copy of one of your datafiles, in which case Oracle will ask you to provide
>all
>the redo log files since then. In a 'normal' case, i.e. shutting down
>'abort',
>killing pmon, unplugging the computer, Oracle is big enough to make sense
>of
>current files. However, when there is a lot of stuff in the current redo
>logs,
>applying them to the datafiles can take some time, which delays the
>availability
>of your database to your users. IMHO, if it was fast in your case it was
>just by
>luck - the synchronisation of memory and datafiles (checkpoint) was not too
>far
>off.
>
>Regards,
>
>Stephane Faroult
>Oriole Corporation
>
>
>
>--
>Please see the official ORACLE-L FAQ: http://www.orafaq.com
>--
>Author: Stephane Faroult
> INET: sfaroult_at_oriole.com
>
>Fat City Network Services -- (858) 538-5051 FAX: (858) 538-5051
>San Diego, California -- Public Internet access / Mailing Lists
>--------------------------------------------------------------------
>To REMOVE yourself from this mailing list, send an E-Mail message
>to: ListGuru_at_fatcity.com (note EXACT spelling of 'ListGuru') and in
>the message BODY, include a line containing: UNSUB ORACLE-L
>(or the name of mailing list you want to be removed from). You may
>also send the HELP command for other information (like subscribing).