Re: Autonomous transactions and deadlocks
From: Hans Forbrich <fuzzy.graybeard_at_gmail.com>
Date: Mon, 25 Mar 2013 06:43:14 -0600
Message-ID: <51504662.2020101_at_gmail.com>
On 25/03/2013 1:56 AM, rajugaru.vij_at_gmail.com wrote:
> Hi,
> Can someone explain me, why does AT cause deadlocks.
>
> Eg: I updated a table x in session 1 and want to update same from session 2 but with AT. It goes into deadlock. But normal update works fine. I am not quite understanding the locking mechanism. Gone through docs, but not clear, or unable to understand. Can some one help please?
I assume you are trying to avoid the mutating table problem while trying to update the same table twice from different transactions. If so, you might describe the requirement in detail to get a reasonable alternative.
Date: Mon, 25 Mar 2013 06:43:14 -0600
Message-ID: <51504662.2020101_at_gmail.com>
On 25/03/2013 1:56 AM, rajugaru.vij_at_gmail.com wrote:
> Hi,
> Can someone explain me, why does AT cause deadlocks.
>
> Eg: I updated a table x in session 1 and want to update same from session 2 but with AT. It goes into deadlock. But normal update works fine. I am not quite understanding the locking mechanism. Gone through docs, but not clear, or unable to understand. Can some one help please?
I assume you are trying to avoid the mutating table problem while trying to update the same table twice from different transactions. If so, you might describe the requirement in detail to get a reasonable alternative.
Depending on he exact problem, you may be tripping over locks on the supporting structures such as constraints or indexes, rather than just locks on the table.
Which version of Oracle Server? New versions have techniques to get around the issue.in triggers.
/Hans
-- http://www.freelists.org/webpage/oracle-lReceived on Mon Mar 25 2013 - 13:43:14 CET