Re: Buffer cache hit very low...!!!
- From: "Michael McMullen" <ganstadba@xxxxxxxxxxx>
- To: <chiragdba@xxxxxxxxx>
- Date: Wed, 14 Sep 2005 08:39:19 -0400
40 Gb database with a 250Mb SGA. What is your db_cache_size? What are you
running on? I would crank that db_cache_size somewhere close to 2Gb and then
never worry about it again. Yes DBCHR is kind of useless if your cache is
already large but it would be valid if your cache is too small. It's not a
magic bullet it will just shut your users up.
As already stated db_sequential_reads means index lookups. Have you tried
analyzing the tables? maybe doing so will give you FTS and improve
performance.
What is the real lesson here? You have users using OEM and telling you
what's wrong. Too many permissions and the tail is wagging the dog. The
users can tell you what's slow but not the reason why.
As an aside. I had a user last week complaining that their query was running
slow all of a sudden. They said it must be oracle because it was using an
index nested loop join. I analyzed the tables and FTS hash join, query went
from not completing ( ie > 12 hrs and cancelled to 8 minutes).
Mike
--
//www.freelists.org/webpage/oracle-l
Other related posts:
