Re: PGA Memory Leak - crashing Server
- From: BN <bnsarma@xxxxxxxxx>
- To: "Alvaro Jose Fernandez" <alvaro.fernandez@xxxxxxxxx>
- Date: Wed, 19 Sep 2007 18:16:34 -0400
Greetings
Its a sort of connection pooling implemented by app.
App opens connections to the database, and keeps them open to service the
requests.
We have never flushed the shared_pool, We didn't think they are connected
and we didn't see any issues towards SP .
Servers got rebooted for running out of Memory because of this issue, in
that process
DB got restarted.
BN
On 9/19/07, Alvaro Jose Fernandez <alvaro.fernandez@xxxxxxxxx> wrote:
>
> Sicne these connections are always there, I wouldn't know if I have
> missed the real issue,
>
> But I did trace 5 sessions and none of them have any thing thats noticable
>
>
>
> one question here. You mean by "connections are always here", that the
> problematic connections are long-lived or something? have you restarted the
> instance or flushed the shared pool
>
> since the problem started?
>
>
>
>
>
> We submitted all the 10046 Level 12 trace files to SR, they came back with
> nothing.
>
>
>
> One thing. are you using OEM/db control or AWR (I'm sorry I'm on 9i yet),
> and maybe you could extract from here some info about the sessions' PGA
> usage, which are accelerating the leakage?
>
>
>
>
>
> thats why I am looking for ideas to
>
>
>
> where to look and ways to reproduce the issues.
>
>
>
> BN
>
>
--
Regards & Thanks
BN
Other related posts:
