Re: RMAN full resync of recovery catalog hangs

From: Steve Harville <steve.harville_at_gmail.com>
Date: Wed, 9 Dec 2009 16:50:07 -0500
Message-ID: <4f41b9050912091350p6ac32583u51bea8519e8d4823_at_mail.gmail.com>

Nothing has been upgraded. Both the target and the repository were created as new 11G databases.
The session in the target database also spends all it's time in SQL*Net message from client waits.
There were some "control file sequential read" at the beginning of the session but those finished quickly.

Steve Harville

http://www.linkedin.com/in/steveharville

On Wed, Dec 9, 2009 at 3:36 PM, Robert Pegram <pegramrg_at_yahoo.com> wrote:

> Steve,
>
> Did you recently run "upgrade catalog"? If so, the first time we upgraded
> our catalog to support 11G databases, we had a couple of target databases
> experiencing similar behavior. Below is my notes from where I resolved the
> problem. Apparently I didn't take good notes, but I seem to remember seeing
> "control file sequential read" events showing up in the targets database's
> v$session_wait view.
>
> =======
>
> Symptom: RMAN full resync of recovery catalog takes extra long time.
>
> FIX:
>
> On Target Database:
> SQL> exec dbms_stats.gather_fixed_objects_stats() ;
> SQL> alter system flush shared_pool;
>
> RMAN> sql "alter session set optimizer_mode=RULE";
> RMAN> resync catalog
>
>
> =======
> HTH
> Rob Pegram
> Duke University
>
>
> ------------------------------
> *From:* Steve Harville <steve.harville_at_gmail.com>
> *To:* Howard Latham <howard.latham_at_gmail.com>
> *Cc:* martin.a.berger_at_gmail.com; oracle-l <oracle-l_at_freelists.org>
> *Sent:* Wed, December 9, 2009 1:21:02 PM
> *Subject:* Re: RMAN full resync of recovery catalog hangs
>
> More Info:
> I'm backing up to tape
> The target and the repository databases are both 11.1.0.7.0
> I have crosschecked and deleted several times, only a few days of backups
> are in the control file.
>
>
> Steve Harville
>
> http://www.linkedin.com/in/steveharville
>
>
> On Wed, Dec 9, 2009 at 11:39 AM, Steve Harville <steve.harville_at_gmail.com>wrote:
>
>> I had done crosscheck and delete several times but I did it again today. I
>> had the same result of resync catalog hanging.
>> I verified that the seq# was the same the entire time while it was waiting
>> on SQL Net message from client.
>>
>> Here's the commands I used:
>>
>> RMAN>crosscheck archivelog all;
>> RMAN>delete noprompt force expired archivelog all;
>>
>> RMAN>crosscheck backup;
>> RMAN>delete noprompt force expired backup;
>>
>> RMAN>crosscheck copy;
>> RMAN>delete noprompt force expired copy;
>>
>>
>>
>> Steve Harville
>>
>> http://www.linkedin.com/in/steveharville
>>
>>
>> On Wed, Dec 9, 2009 at 2:57 AM, Howard Latham <howard.latham_at_gmail.com>wrote:
>>
>>> I had a similar problem after several attempts at crosscheck and delete
>>> expired it all started working again.
>>>
>>> 2009/12/9 Martin Berger <martin.a.berger_at_gmail.com>
>>>
>>> Steve,
>>>>
>>>> just to make sure: is the seq# the same all the time for the session
>>>> with " waiting for SQL*Net" event?
>>>>
>>>> Martin
>>>>
>>>>
>>>> On Tue, Dec 8, 2009 at 15:22, Steve Harville <steve.harville_at_gmail.com
>>>> > wrote:
>>>>
>>>>> RMAN full resync of recovery catalog hangs
>>>>>
>>>>> When I check in the catalog database the session is waiting for SQL*Net
>>>>> message from client.
>>>>> ....
>>>>>
>>>>
>>>>
>>>>
>>>
>>>
>>>
>>> --
>>> Howard A. Latham
>>>
>>>
>>>
>>
>
>

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

Received on Wed Dec 09 2009 - 15:50:07 CST