Argh.....
- From: "Bobak, Mark" <Mark.Bobak@xxxxxxxxxxxxxxx>
- To: <oracle-l@xxxxxxxxxxxxx>
- Date: Wed, 15 Sep 2004 16:45:15 -0400
We are suffering from a bunch of excess database calls, resulting in
lots more executions of SQL and (ovbiously) more buffer gets for any
given interval of time, and of course, cache buffers chains latch
contention is the result.
So, what's my boss's direction? Find the source of the extra calls and
extra workload? No.....
"Mark, when we moved to 9i, you removed _db_block_hash_latches and
_db_block_hash_buckets right?"
"Yeah, well, they're undocumented, I couldn't really see a justification
for keeping them, and with such a major upgrade (8.1.7.4. to 9.2.0.5), I
thought it made sense to start out w/ a baseline of not having them
set."
(This is my favorite part...)
"I feel that is the problem....Please look into setting these parameters
again...."
("I feel..." Grrr.....it just gets on my nerves...."I feel". How about,
"Based on this data, I think this is what's happening"?? "I feel"?? Or
"Here's the evidence which suggests this is the problem....")
Ok, I'm going to go now and pretend that setting these parameters is
going to have any significant effect on solving the problem.
--
Mark J. Bobak
Oracle DBA
ProQuest Company
Ann Arbor, MI
"On two occasions, I have been asked [by members of Parliament], "Pray,
Mr. Babbage, if you put into the machine wrong figures, will the right
answers come out?' I am not able to rightly apprehend the kind of
confusion of ideas that could provoke such a question."
-- Charles Babbage (1791-1871)
--
//www.freelists.org/webpage/oracle-l
Other related posts:
