GeoPappas wrote:
> Background: I created a brand new (from scratch) 10gR1 instance using
> ASM on Solaris 2.9 back in 07/15/2005. I then upgraded the instance to
> 10gR2 (10.2.0.1) about 4 months later (on 11/11/2005).
>
> I have recently noticed that I am receiving "Thread 1 cannot allocate"
> warning messages in my alert log every now and then.
>
> The puzzling thing is that these messages started immediately after the
> upgrade to 10gR2 and have continued every now and then since then.
> That is, there were no messages before the 10gR2 upgrade (for about 4
> months).
>
> My question is: What is different about 10gR2 that it would cause these
> errors to surface???
>
> Other useful info:
>
> - There are a total of 20 messages since 11/11/2005, so they are not
> that frequent.
> - I currently have 6 redo logs and they are each 25 MB. They normally
> switch every few hours.
> - The database instance uses an ASM instance.
> - The load after the 10gR2 upgrade is about the same as before the
> upgrade. That is, there hasn't been a large increase in load since the
> upgrade.
I had a problem with one database I upgraded where Oracle did not change
the version number in the log files causing a mismatch. Unfortunately I
can not longer recall the error message I received and you didn't post
yours. You might want to try dropping one log file group and recreating
it and then the rest of them, sequentially, if no other solution
presents itself.
--
Daniel A. Morgan
http://www.psoug.org
damorgan_at_x.washington.edu
(replace x with u to respond)
Received on Thu Jan 05 2006 - 11:44:56 CST
