Re: oracle 9.0.1.5 windows handle leak ?
From: Mark D Powell <Mark.Powell_at_eds.com>
Date: Tue, 15 Sep 2009 08:27:28 -0700 (PDT)
Message-ID: <9f52a24f-c2e6-4b49-8706-343f972880ce_at_y20g2000vbk.googlegroups.com>
On Sep 12, 12:58 pm, mokk <i..._at_mail.ru> wrote:
> Thanks for your response
>
> Maybe I should point out that I use one listener for both db , of
> course version 10.2.
> Further experiments leads to more serious non-paged pool leak in
> tnslsnr.exe.And I found in metalink the problem
> Doc ID: 472181.1
> Non-Paged Pool Memory and Handle Count Leak While Using 10.2 Listener
> for 10.1 and 9i Databases
> Setting USE_SHARED_SOCKET seems solved it.
> It's interesting that the problem exists even if 10.2 and 9.0.1
> databases reside on different machines and 9.0.1 db use listener of
> its own version.And setting USE_SHARED_SOCKET also helps .
Date: Tue, 15 Sep 2009 08:27:28 -0700 (PDT)
Message-ID: <9f52a24f-c2e6-4b49-8706-343f972880ce_at_y20g2000vbk.googlegroups.com>
On Sep 12, 12:58 pm, mokk <i..._at_mail.ru> wrote:
> Thanks for your response
>
> Maybe I should point out that I use one listener for both db , of
> course version 10.2.
> Further experiments leads to more serious non-paged pool leak in
> tnslsnr.exe.And I found in metalink the problem
> Doc ID: 472181.1
> Non-Paged Pool Memory and Handle Count Leak While Using 10.2 Listener
> for 10.1 and 9i Databases
> Setting USE_SHARED_SOCKET seems solved it.
> It's interesting that the problem exists even if 10.2 and 9.0.1
> databases reside on different machines and 9.0.1 db use listener of
> its own version.And setting USE_SHARED_SOCKET also helps .
Glad you found the correct note for your issue.
- Mark --