After just migrating from 8.0.6 to 8.1.7 and RBO to CBO, I've been
researching about perf tuning and have come across the init.ora parameter
OPTIMIZER_INDEX_COST_ADJ. The guide, by Tim Gorman of evdbt.com, explained
that it's default value is 100 (percent), but should probably be in the
10-50 range for OLTP systems. It recommended using the following query as a
guideline to set this:
SELECT event, average_wait
FROM v$system_event
WHERE event LIKE 'db file s%read';
...and to take the ratio of "db file sequential read" to "db file scattered
read" as a value for OPTIMIZER_INDEX_COST_ADJ. That may be fine for most
systems, but our ratio is around .08%. No, it's not 8% -- and it's not a
decimal problem. I'm getting this:
db file sequential read 1.83687542789106
db file scattered read 2151.12743289383
I'm reasonably certain that the seemingly outrageous "2151.127" is due to
our use of HP's AutoRAID (SLOWWWWWW!) -- we won't be getting one in our next
box.
So, I'm a little skiddish about changing the default value from 100 to 1 or
even 10, based on the above query alone. The scattered read average wait
seems to be dropping steadily (it's now down to ~1900 over a period of 14
hours), so I'm trying to provide a "good" value for this parm without
adversely affecting the optimizer.
Anyone have any ideas?
TIA!
Rich Jesse System/Database Administrator
[EMAIL PROTECTED] Quad/Tech International, Sussex, WI USA
--
Please see the official ORACLE-L FAQ: http://www.orafaq.com
--
Author: Jesse, Rich
INET: [EMAIL PROTECTED]
Fat City Network Services -- (858) 538-5051 FAX: (858) 538-5051
San Diego, California -- Public Internet access / Mailing Lists
--------------------------------------------------------------------
To REMOVE yourself from this mailing list, send an E-Mail message
to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in
the message BODY, include a line containing: UNSUB ORACLE-L
(or the name of mailing list you want to be removed from). You may
also send the HELP command for other information (like subscribing).
