RE: COMMIT as a long running text

It's not a big deal to get this from v$session.sql_hash_value using last_et_time as criteria. I think this effect is due to pl/sql block execution. last_et_time time indicates a long operation while IMHO v$session.sql_hash_value is just changing for every sql statement in the the pl/ssql block, or something like that.

-----Original Message-----
From: oracle-l-bounce_at_freelists.org [mailto:oracle-l-bounce_at_freelists.org] On Behalf Of rjamya Sent: 12. október 2006 11:17
To: tanmoydc_at_gmail.com
Cc: oracle-l_at_freelists.org
Subject: Re: COMMIT as a long running text

what tool? what is the hash value of those statements? zero? if so then some tools (e.g. Confio Ignite) treat them as " DDL or commit ", but actually those could be dbms_jobs, in addition to commits.

Raj

On 10/11/06, Tanmoy <tanmoydc_at_gmail.com> wrote:
>
> We have some monitoring tools like long running query for 2 hrs etc ,
> One of our long running query alert was showing a customized
> application having sql text COMMIT as running more than 2hrs.
> We found out mutiple session sometimes having the same sql text.
> considering considerable IO on server.

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


Fyrirvari/Disclaimer
http://www.landsbanki.is/disclaimer
--
http://www.freelists.org/webpage/oracle-l