Re: Deadlocks

From: chet justice <chet.justice_at_gmail.com>
Date: Tue, 29 Sep 2009 21:49:10 -0400
Message-ID: <8311a5b60909291849i59cb72dcgd62a7d917cf8faca_at_mail.gmail.com>

Miladin Modrakovic (aka Oraclue <http://oraclue.com>), has 2 pretty good articles on it.

Part 1: http://oraclue.com/2009/04/20/detecting-deadlock-source/

Part 2: http://oraclue.com/2009/04/23/detecting-deadlock-source-part-2/

I worked with him when we began receiving those deadlocks (me in a junior/dev dba role). My contention was that it was due to a terrible design (cascading triggers and such) but, I believe, he ultimately solved the problem by indexing unindexed foreign keys. He did some pretty in-depth research on it and those articles are just a bit of what he found.

chet

On Tue, Sep 29, 2009 at 4:12 PM, Alex Fatkulin <afatkulin_at_gmail.com> wrote:

> Oracle never kills the session as a result of ora-60. The session
> which received ora-60 can decide whether commit the work which was
> already done or repeat/modify/skip the statement.
>
> On Tue, Sep 29, 2009 at 4:03 PM, Guillermo Alan Bort
> <cicciuxdba_at_gmail.com> wrote:
> > If I recall correctly, a deadlock gives an ora-60 error in the alert log
> and
> > kills one of the two sessions...
>
> > hth
> > Alan Bort
> > Oracle Certified Professional
> >
>
>
>
> --
> Alex Fatkulin,
> http://afatkulin.blogspot.com
> http://www.linkedin.com/in/alexfatkulin
> --
> http://www.freelists.org/webpage/oracle-l
>
>
>

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

Received on Tue Sep 29 2009 - 20:49:10 CDT