RE: 'buffer busy waits' on Header Block (#2) of Tempfile

Since you're apparently competing for allocation within a single temporary tablespace, it is likely that one of the variety of having multiple temporary tablespaces will help you.

Adding multiple tablespaces to a tablespace group and using the tablespace group for the temporary tablespace is the probably the simplest thing to try first if you're up to a release that supports it. (That's in the ALTER USER and ALTER tablespace bits of the documentation). If you're using a BORING allocation rather than SAME, you might put the temporary tablespaces that are components of the tablespace group on different stripesets to fan out the i/o.

If either that does not reduce the wait or you're not up to that level, then you could consider creating multiple tablespaces and statistically assigning the temporary tablespace to be used in a logon trigger to load level your users, or just use a fixed distribution of users to tablespaces if that is an even enough distribution of users per temporary tablespace at run time.

Of course if you have sufficient memory to reduce the need to use temporary tablespace i/o you could try that. Often a pretty reasonable upward bump in PGA_AGGREGATE_TARGET dramatically reduces i/o.

Another thing is to examine your extent size. A large uniform extent size will tend to mean fewer allocations take place. Usually disk acreage for temp is not that much of a concern, since disk SPACE is cheap and you don't have to back up temp space.

I'd probably evaluate using additional memory first, but if you have the acreage available there is little or no down side to adding more tablespaces.

Good luck!

-----Original Message-----
From: oracle-l-bounce_at_freelists.org [mailto:oracle-l-bounce_at_freelists.org] On Behalf Of Hemant K Chitale
Sent: Monday, April 28, 2008 8:36 AM
To: oracle-l_at_freelists.org
Subject: 'buffer busy waits' on Header Block (#2) of Tempfile

We'd rolled out a new system today and user connections rapidly went up to 900 sessions. OK, we are supposed to be able to handle that.

However, by the afternoon, I saw 'buffer busy waits' on Block#2 of the first tempfile of the Temporary Tablespace (using a custom Temporary tablespace instead of 'TEMP', although the default 'TEMP' exists). By evening, we had 300 sessions waiting on 'buffer busy waits' on the same block
(querying V$SESSION_WAIT for P1, P2).

Although the tablespace has 3 tempfiles, the first file has 350 sessions against it and the other two have less than 100 sesssions put together
(querying V$TEMPSEG_USAGE for SEGFILE#).
These are a mix of SORT and HASH extents.

How can I address this ?

1. Increase PGA_AGGREGATE_TARGET (1GB for 400 concurrent users + 10-15 batch jobs) and/or switch to WORKAREA_SIZE_POLICY='MANUAL' with SORT_AREA_SIZE and HASH_AREA_SIZE.
2. Add more tempfiles to the tablespace and/or rebuild the tablespace with larger Extent Sizes (1MB instead of the current value of 256K)

Any other suggestions ?

Hemant K Chitale
http://hemantoracledba.blogspot.com

"A 'No' uttered from the deepest conviction is better than a 'Yes' merely uttered to please, or worse, to avoid trouble." Mohandas Gandhi Quotes
: http://www.brainyquote.com/quotes/authors/m/mohandas_gandhi.html

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




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