Comments on: Obsolete cursors

By: 11.2: ожидание cursor: pin S wait on X при избыточной генерации курсоров параллельного выполнения | Oracle mechanics

Sat, 02 Mar 2013 10:36:31 +0000

[...] – интересным было только ярко выраженное верхнее ограничение количества дочерних курсоров на запрос, определяемое параметром _cursor_obsolete_threshold – см. Timur Akhmadeev. Obsolete cursors: [...]

By: Masaaki Yoshikawa

Masaaki Yoshikawa — Tue, 25 Dec 2012 06:25:01 +0000

Happy Holidays! Timur-san,

I applied 14127231 one-off patch.
However, the problem of remaining V$SQL.IS_OBSOLETE=’Y’ records occurred.
Sad.;-<
I want to continue my SR.

By: Masaaki Yoshikawa

Masaaki Yoshikawa — Fri, 21 Dec 2012 06:58:46 +0000

Thank you for good information! I will confirm the details later.

By: Timur Akhmadeev

Timur Akhmadeev — Fri, 21 Dec 2012 06:36:04 +0000

One more bug related to obsolete cursors in 11.2.0.3: Bug 14127231 dbms_shared_pool.purge raised ora-6570 on obsoleted child cursors. Fixed in 11.2.0.4, one-off patches exists for 11.2.0.3 for some platforms. This bug-fix could address the real issue (obsolete v$sql entries) - I didn't test it though.

By: Timur Akhmadeev

Timur Akhmadeev — Thu, 13 Dec 2012 06:57:18 +0000

Hello Masaaki-san

thank you very much for your consistency!

By: Masaaki Yoshikawa

Masaaki Yoshikawa — Thu, 13 Dec 2012 04:56:04 +0000

Hello Timur-san,

I understood my argument was off the point.

My customer sometimes got a huge garbage plan and was troubled.
Therefore I hurried a fix of ORA-1422 error and have forgotten a root problem.

Now I intend to register SR which assigned a focus to a problem of remaining V$SQL.IS_OBSOLETE=’N’ records.
Thank you for good advice.

Masaaki Yoshikawa(Mr.)

By: Timur Akhmadeev

Timur Akhmadeev — Wed, 12 Dec 2012 09:54:36 +0000

Hi Masaaki-san

Thank you for the update. I’ve downloaded the one-off patch and discovered that they fix DBMS_XPLAN code, which will fix the exception, but the fix to the package is not what I would like to see in the patch. They are fixing consequences and not the root cause. They have to fix V$SQL to exclude these obsoleted entries (this is exception root cause), otherwise there’ll be more bugs coming.
Sad.

By: Masaaki Yoshikawa

Masaaki Yoshikawa — Tue, 11 Dec 2012 05:54:34 +0000

Hi Timur-san,
Additional information.
Individual patch 14585499 for 11.2.0.3.X Generic was shown, now.

Ciao,

By: Timur Akhmadeev

Timur Akhmadeev — Mon, 29 Oct 2012 07:37:31 +0000

Dear Masaaki-san

thanks a lot for the bug number.

By: Masaaki Yoshikawa

Masaaki Yoshikawa — Mon, 29 Oct 2012 05:04:53 +0000

Dear Timur-san,

I tried the solution of this symptom.
I wanna thank you so much for your good testcase.
This problem was fixed.
BUG:14585499 PSR11.2.0.4

By: Excess Child Cursors and Cardinality Feedback | Mike Pargeter

Excess Child Cursors and Cardinality Feedback | Mike Pargeter — Tue, 24 Jul 2012 15:58:39 +0000

[...] there was an even distribution of 23 or 24 rows per CHILD_NUMBER. I read about a similar situation here. In summary, 11gR2 has a limit of 100 children per statement, but also keeps invalidated rows in [...]

By: Flashback cursors « Timur Akhmadeev's blog

Flashback cursors « Timur Akhmadeev's blog — Wed, 06 Jun 2012 09:14:59 +0000

[...] it causes hundreds of child cursors to appear in Oracle versions before 11.2.0.3 (this version limits number of child cursors to 100 by default). Of course they should be purged from shared pool [...]