Re: direct path write wait
- From: "Paul Drake" <bdbafh@xxxxxxxxx>
- To: deepak_blr@xxxxxxxxxxx
- Date: Wed, 27 Dec 2006 11:02:12 -0500
On 12/27/06, blr_dba <deepak_blr@xxxxxxxxxxx> wrote:
Dear Gurus,
Am having huge amount of waits on "direct path write event". This event
has
98% of the total wait time. Now we want to tune it.
Perhaps you might first want to tune the SQL statements that are causing the
writes to tempfiles (sorts, hash_joins, GTT usage)?
Pd
We are using Solaris 9 and Oracle 9.2.0.7.
After examining the system we found that logging is enabled on the mounted
UFS file systems.
Need your expert advice on the following...
> How to know whether the UFS is enabled for asynchronous IO or direct IO?
> Which one is better for DWH environments? Pros and cons of each.
> Is filesystem cache is involved in both of the above methods?
Please suggest the best solution for my environment.
Looking forward to your ever valuable feedback.
Deepak
--
http://www.freelists.org/webpage/oracle-l
--
-- ALTER SESSION SET EVENTS 'immediate trace name hanganalyze level 4';
- Follow-Ups:
- Re: direct path write wait
- From: blr_dba
- References:
- RE: AIX pinned memory
- From: Adrian
- direct path write wait
- From: blr_dba
Other related posts:
- » direct path write wait
- » Re: direct path write wait
- » Re: direct path write wait
- » Re: direct path write wait
- » Re: direct path write wait
- » Re: direct path write wait
Dear Gurus, Am having huge amount of waits on "direct path write event". This event has 98% of the total wait time. Now we want to tune it.
After examining the system we found that logging is enabled on the mounted UFS file systems. Need your expert advice on the following... > How to know whether the UFS is enabled for asynchronous IO or direct IO? > Which one is better for DWH environments? Pros and cons of each. > Is filesystem cache is involved in both of the above methods? Please suggest the best solution for my environment. Looking forward to your ever valuable feedback. Deepak -- http://www.freelists.org/webpage/oracle-l
- Re: direct path write wait
- From: blr_dba
- RE: AIX pinned memory
- From: Adrian
- direct path write wait
- From: blr_dba