RE: Shutdown Abort

From: <Joel.Patterson_at_crowley.com>
Date: Tue, 3 Jul 2007 16:12:24 -0400
Message-ID: <02C2FA1C9961934BB6D16DE35707B27B031C2AFD@jax-mbh-01.jax.crowley.com>

Remember, this was directed at my observation. I am not advocating, I am being a messenger in this case. I have done a lot of aborts, without consequence, (and if my memory serves me back to Oracle 7, some with consequence - but don't remember specifics).

So, maybe it was coincidence. Don't take the observed, and mailed message as an argument against abort.

That's not to say there wasn't something else wrong as has been indicated. However, I doubt I will be doing any 'testing' on this as was also stated - there are bigger fish to fry, and this database is up 24/7 with RMAN hot backups. I have a provision in my rman shell script, where at the end, I can shutdown 'immediate' and startup to execute special commands automagically while I sleep. Most often to create a new spfile.

Why is it immediate? Well.... I don't know that abort was necessary
(even if it is safe). Immediate doesn't take very long on our systems,
I'm sleeping anyway, and it just seemed like a nicer way to shutdown/startup while I'm sleeping. So there isn't anything really 'wrong' about immediate is there?

Shutdown immediate has been a matter of course since Oracle 7, because shutdown normal is - well not going to happen. Shutdown abort was always done interactively if needed. That great of an availability SLA wasn't a big issue unless there was an abnormal downed production system during business hours. Just, most of the time, it seems everyone grew up with immediate -- (again just a small observation based on an insignificant number of DBA's -- say 15 over a decade).

Should I change my script to 'abort' in order to create a new spfile?
(or whatever task I program in)? So far the argument is that abort is
not a bad thing, but indeed saves a lot of time, which can be regained during startup. What if that is not an issue?

The impression is beginning to look like as SOP, 'abort' has become the de-facto standard over immediate. I do wonder why oracle isn't more explicit with this parameter in it's documentation. - but before you kill me on that, I am sure I am not as familiar with the Docs as you guys, so if they have, be nice.

Perhaps you will convince me that abort should be SOP (as opposed to not bad).... But then were does that leave immediate?

Going home now, be back tomorrow.

Joel Patterson
Database Administrator
joel.patterson_at_crowley.com
x72546
904 727-2546

From: oracle-l-bounce_at_freelists.org
[mailto:oracle-l-bounce_at_freelists.org] On Behalf Of Robert Freeman Sent: Tuesday, July 03, 2007 1:54 PM
To: jkstill_at_gmail.com
Cc: oracle-l_at_freelists.org
Subject: RE: Shutdown Abort

Old arguments never die, they just get new proponents...

Robert G. Freeman
Oracle Consultant/DBA/Author
Principal Engineer/Team Manager
The Church of Jesus Christ of Latter-Day Saints Father of Five, Husband of One,
Author of various geeky computer titles
from Osborne/McGraw Hill (Oracle Press)
Oracle Database 11g New Features Now Available for Pre-sales on Amazon.com!
Sig V1.1

	-----Original Message-----
	From: oracle-l-bounce_at_freelists.org
[mailto:oracle-l-bounce_at_freelists.org]On Behalf Of Jared Still
	Sent: Tuesday, July 03, 2007 10:23 AM
	To: randyjo_at_sbcglobal.net
	Cc: oracle-l_at_freelists.org
	Subject: Re: Shutdown Abort

	On 7/3/07, Randy Johnson <randyjo_at_sbcglobal.net> wrote: 

		With all due respect this was not a controlled test. So

I'd be carefull not to assume it was caused by the abort.

Exactly.

Please see
http://www.speakeasy.org/~jwilton/oracle/shutdown-abort-bad.html

If you are not familiar with Jeremiah Wilton, he was a DBA at Amazon.com from early days.

Amazon has a few databases, and they were/are regularly shutdown with abort.

	Jared Still
	Certifiable Oracle DBA and Part Time Perl Evangelist

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

Received on Tue Jul 03 2007 - 15:12:24 CDT