Re: Oracle 10g Rel.2 taking longer time when SHUTDOWN IMMEDIATE

I second Niall's questions:

• Why "bounce" every weekend, or on any regular schedule?
• What is wrong with SHUTDOWN ABORT, then STARTUP RESTRICT (a.k.a. STARTUP FORCE RESTRICT), then SHUTDOWN IMMEDIATE as a normal procedure for getting the database down?

As far as answering your question directly, have you tried enabling ALTER SYSTEM SET SQL_TRACE = TRUE immediately prior to issuing SHUTDOWN IMMEDIATE? Yes, I suggested ALTER SYSTEM SET, not SESSION, because it might be an unrelated Oracle processes that is performing the work that is delaying shutdown. If the culprit isn't the session that is issuing the SHUTDOWN IMMEDIATE itself, then surely it will be either PMON or SMON. As far as being squeamish about turning SQL tracing on for the entire instance, what can it hurt if you have enough disk space in the USER/BACKGROUND_DUMP_DEST directories?

And think of what you'll learn... :-)

...and please feel free to post what you learn back to the list? :-)

Niall Litchfield wrote:
> A couple of questions.
>
> Why bounce the databases at all if downtime of a few tens of minutes
> matters? My general position is that where a database is being
> deliberately taken down on a schedule that isn't for patching etc
> something (app/backup strategy etc) is wrong. There could be a good
> reason for it, but often there isn't.
>
> If you must bounce the db what is wrong with shutdown abort? You don't
> look as if this is for a cold backup so who cares that some recovery
> is required?
>
> On 8/11/06, * Syed Jaffar Hussain* <sjaffarhussain_at_gmail.com
> <mailto:sjaffarhussain_at_gmail.com>> wrote:
>
> Hello List,
>
> Since migrated from 9i to 10g, we are facing few new issues.
> At our premises, we tend to bounce our database at week-end. In 9i, it
> was talking few minutes. But, in 10g Re2.10, it is taking more 20
> minutes.
> When we shutdown immediate, there would be around 600+ inactive
> sessions. In 9i, there was not issues with the number of inactive
> session, but, in 10g Rel.2, I believe, some internal changes have
> been
> made while killing the inactive session. Thats why, it might be taking
> longer time.
> I am afraid, if this is the case, in case of emergency, we need to
> wait 20+ minutes to bring down the database. I dont want to opt for
> 'shutdown abort' option.
> Is anyone had/having the same issue with 'shutdown immediate' in
> Oracle 10g Rel.2?
>
>
> --
> Best Regards,
> Syed Jaffar Hussain
> 8i,9i & 10g OCP DBA
> Banque Saudi Fransi,
> Saudi Arabia
>
> I blog at :http://jaffardba.blogspot.com/
> http://www.oracle.com/technology/community/oracle_ace/ace1.html#hussain
> <http://www.oracle.com/technology/community/oracle_ace/ace1.html#hussain>
> ----------------------------------------------------------------------------------
> "Winners don't do different things. They do things differently."
> --
> http://www.freelists.org/webpage/oracle-l
>
>
>
>
>
> --
> Niall Litchfield
> Oracle DBA
> http://www.orawin.info

--
http://www.freelists.org/webpage/oracle-l