Re: x$ksmlru
- From: Jay Hostetter <hostetter.jay@xxxxxxxxx>
- To: oracle-l@xxxxxxxxxxxxx
- Date: Tue, 20 Jul 2010 08:28:28 -0400
Just thought I'd post this follow-up to my recurring ORA-4031 issue. I
have several suspect queries that will need further investigation. Howver,
I may have found the main culprit - I was trending my SGA stats and noticed
that "miscellaneous" continues to grow over time (pivot charts in Excel 2007
are great!). This happens to be a known bug in Oracle 9.2.0.4 (see Note
316138.1) that causes Ora-4031 errors. Hopefully the workaround will help
to alleviate the issue.
Thanks again for the responses.
Jay
> Mark and Tanel,
>
>
>
> Thank you for the responses. I suspect a regularly scheduled "bad"
> process as well.
>
> I notice that these statements with the 26284 unsharable size occur
> several times throughout the week - not just on the day of the 4031 errors.
> The KSMLRHON and KSMLROHV columns contain no information that would lead me
> back to the culprit. I've been monitoring V$SQL for statements that have
> high runtime_mem. Is there a better way to capture the problem statement
> (other than heapdumps - I'm afraid this database is too volatile)?
>
> I've been reading Tanel's blog posts on memory troublshooting. I've able
> to query x$ksmsp without any repurcussions, but this hasn't told me what is
> causing the memory issues.
>
>
Other related posts: