| Oracle FAQ | Your Portal to the Oracle Knowledge Grid | |
 |
 | |||
| Oracle FAQ | Your Portal to the Oracle Knowledge Grid | |
|
| |||
Home -> Community -> Mailing Lists -> Oracle-L -> RE: The best CPU usage measurement in Oracle: BUFFER_GETS or CPU_TIME?
Cary,
I just wonder are there situation, then the usage of SQL list sorted by
BUFFER_GETS (as it is now in statspack) is more useful then sorted by
CTU_TIME?
If NO, then why there still exist TOP sorted by BUFFERS?
Can you imagine the situation then BUFFER_GETS is better then CPU_TIME? Then I need to look on BUFFER_GETS first of all?
Jurijs
PS
As you wrote in your book, CPU_TIME can be not so reliable statistic
(measurement cs on the most platforms, double counting etc.). But it is
seams CPU_TIME statistics reflect situation more accurate even with
measurement errors, then BUFFER_GETS
Jurijs
9268222
"Cary Millsap" <cary.millsap_at_hotsos.com>
Sent by: oracle-l-bounce_at_freelists.org
21.06.2004 17:55
Please respond to oracle-l
To: <oracle-l_at_freelists.org>
cc:
Subject: RE: The best CPU usage measurement in Oracle:
BUFFER_GETS or CPU_TIME?
Jurijs,
If, in step 1, you've discovered that either (a) 'CPU service' or (b)
competition for CPU service (what the book calls 'unaccounted for' time) =
is
the major source of your response time consumption, then the right step =
2 is
to identify the top consumers of CPU_TIME during the problem time =
interval.
The direct statistic, not an indirect or incomplete one like BUFFER_GETS =
for
exactly the reasons you have described.
This is true /if/ CPU_TIME is a reliable statistic in Oracle. I have not investigated whether it is, but I do expect that it is.
Cary Millsap
Hotsos Enterprises, Ltd.
http://www.hotsos.com
* Nullius in verba *
Upcoming events:
- Performance Diagnosis 101: 6/22 Pittsburgh, 7/20 Cleveland, 8/10 =
Boston
- SQL Optimization 101: 5/24 San Diego, 6/14 Chicago, 6/28 Denver - Hotsos Symposium 2005: March 6-10 Dallas - Visit www.hotsos.com for schedule details...
-----Original Message-----
From: oracle-l-bounce_at_freelists.org =
[mailto:oracle-l-bounce_at_freelists.org]
On Behalf Of J.Velikanovs_at_alise.lv
Sent: Monday, June 21, 2004 3:05 AM
To: oracle-l_at_freelists.org
Subject: RE: The best CPU usage measurement in Oracle: BUFFER_GETS or
CPU_TIME?
>> My point is: may be now the better measure is CPU_TIME? I show in=20
previous
>> posting situation then CPU_TIME figure is 10 times accurate then=20
BUFFER_GETS.
>> Can you show me contrariwise situation?=20
>>=20
>> [Cary Millsap] No, I think you're actually on the best path to the=20
solution.
Lets me ask contrariwise question:
Are there situations when using statspack report (if using) or other=20
instance wide performance tool the better way to see TOP SQL-s is sort =
by=20
BUFFER_GETS not CPU_TIME?
Jurijs
Additional info from previous postings:
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
1. Activity=20
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
NAME VALUE
----------------------------------------------------------------=20session logical reads 23002
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
2. Activity=20
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
NAME VALUE
----------------------------------------------------------------=20session logical reads 23006
I.
session logical reads=20
Action 1 =3D Action 2
II.
CPU used by this session
Action 1 * 10 =3D Action 2
-- Archives are at http://www.freelists.org/archives/oracle-l/ FAQ is at http://www.freelists.org/help/fom-serve/cache/1.html ----------------------------------------------------------------- ---------------------------------------------------------------- Please see the official ORACLE-L FAQ: http://www.orafaq.com ---------------------------------------------------------------- To unsubscribe send email to: oracle-l-request_at_freelists.org put 'unsubscribe' in the subject line. -- Archives are at http://www.freelists.org/archives/oracle-l/ FAQ is at http://www.freelists.org/help/fom-serve/cache/1.html ----------------------------------------------------------------- ---------------------------------------------------------------- Please see the official ORACLE-L FAQ: http://www.orafaq.com ---------------------------------------------------------------- To unsubscribe send email to: oracle-l-request_at_freelists.org put 'unsubscribe' in the subject line. -- Archives are at http://www.freelists.org/archives/oracle-l/ FAQ is at http://www.freelists.org/help/fom-serve/cache/1.html -----------------------------------------------------------------Received on Mon Jun 21 2004 - 10:10:09 CDT
 |
 |