It's 800 sec for 5617 exec/fetch calls. It looks like it's a sql inside a
cursor loop or stored proc that gets called from some app.
If you are unhappy, try to get rid of the cursor logic and get everything
done in one sql call.
Waleed
-Original Message-
Sent: Tuesday, October 28,
Have you run SQLTRACE on this query?
The detail in the trace file will show
where the cpu is being consumed.
There is insufficient data in the summary
to reach any conclusion.
Jared
On Mon, 2003-10-27 at 21:24, Muqthar Ahmed wrote:
Hi,
Execution plan looks good but the query is consuming
Jared,
The output is from TKPROF.
Muqthar
-Original Message-
Sent: Tuesday, October 28, 2003 9:24 AM
To: Multiple recipients of list ORACLE-L
Have you run SQLTRACE on this query?
The detail in the trace file will show
where the cpu is being consumed.
There is insufficient data in
just in case, if you have specified explain= on the command line to generate tkprof
summary, remove it and just do
tkprof filename ...
which version of oracle?
Raj
Rajendra dot Jamadagni at nospamespn dot com
All
is good but HIGH CPU
Jared,
The output is from TKPROF.
Muqthar
-Original Message-
Sent: Tuesday, October 28, 2003 9:24 AM
To: Multiple recipients of list ORACLE-L
Have you run SQLTRACE on this query?
The detail in the trace file will show
where the cpu is being consumed
Hi,
The developer was executing the SQL statement in the loop in procedure. After
re-writing the code, now the performance is good.
Thanks
Muqthar Ahmed
-Original Message-
Sent: Tuesday, October 28, 2003 8:09 AM
To: Multiple recipients of list ORACLE-L
It's 800 sec for 5617
]
cc:
Subject:RE: Execution Plan is good but HIGH CPU
Jared,
The output is from TKPROF.
Muqthar
-Original Message-
Sent: Tuesday, October 28, 2003 9:24 AM
To: Multiple recipients of list ORACLE-L
Have you run SQLTRACE on this query?
The detail in the trace file will show
Hi,
Execution plan looks good but the query is consuming 800 seconds CPU timewhy?
SELECT sampleavail, sample_cost_amount, sample_sale_amount,
discount_room, discount_case, discount_half_case, allow_cut,
Hi,
I guess you are just using RBO and did not analyze your table. Try analyze it and
run it again.
You nested loop maybe is inefficient, as it generate a lot of buffer_gets.Maybe
you can consider using hash_join instead of nested loop.
If you still plan to use nested loop, consider
OS :hp-ux
oracle rdbms 8.1.7.4
there are certain processes taking lot of cpu (
some thing went wrong with stastics last nite). which is causing almost a halt
in production database.these processes are killed from oracle but i can still
see them on OS .Cpu is still high .
I though after
Rebounce the database...
After shutting down database , check for any oracle process is still alive
by using
ipcs -mb
If it is still their kill those process by yourself (using ipcem -m ) or by
Unix admin. Sometime it does not go and then you have to reboot.
If shutdown immediate takes
Ak - Were the processes you killed doing updates? If so the work is probably
rolling back changes. Then it won't make much difference whether the process
rolls it back or smon does. You could take a look at the wait statistics to
get an idea of what is going on.
Dennis Williams
DBA, 80%OCP,
, 100% DBA
Lifetouch, Inc.
[EMAIL PROTECTED]
-Original Message-
From: AK [mailto:[EMAIL PROTECTED]
Sent: Wednesday, June 11, 2003 5:04 PM
To: Multiple recipients of list ORACLE-L
Subject: high cpu
OS :hp-ux
oracle rdbms 8.1.7.4
there are certain processes taking lot of cpu ( some
why dont you check the v$transaction to see whether it is rolling back
or i mean it is still active.
Sometimes if a process is doing lots of reads on the database and causing lots of cpu spike it would be better you kill the process from the database first and then from the side unix also,we had
Message-
From: Seema Singh [mailto:oracledbam;hotmail.com]
Sent: Monday, November 04, 2002 2:09 PM
To: Multiple recipients of list ORACLE-L
Subject: MTS process HIGH CPU
Hi
I am using MTS in 8.1.6.I am wondering one of MTS process is
locked 2 table
since morning and this process
Hi
I am using MTS in 8.1.6.I am wondering one of MTS process is locked 2 table
since morning and this process is taking more CPU.Can I kill that process
like ora_s000_prod by using alter system and then unix kill command?
Thx
-Seema
Guys,
The query was rewritten with only one thing in mind - Logical IO. As
suggested, I was able to recode the query and what a difference it was. The
cpu utilization is now only 10% of the same query executing before. Thanks
all for the timely help. Infact, when Farrukh suggested the index is
I've got a little lost about who has said what in answer to
whom about what - so apologies if I'm repeating comments,
answering non-questions and giving incorrect attributions.
The primary problem appears to be that performance
plunges dramatically when concurrent increases.
Three tkprof
Again: needed time is 1.76
20 jobs require 1.76 * 20
Divided on 4 cpu
each cpu will take 1.76 * 20 /4 = 8.8 sec
This is the average elapsed time for any job (20 concurrent) on your
system.
regards,
Waleed
-Original Message-
To: Multiple recipients of list ORACLE-L
Sent:
The elapsed time taken can not change for this stress test except by
improving the cpu time needed to execute a single task.
So to improve the numbers some tuning needed on the sql.
The average elapsed time for (n concurrent jobs) = x * n / c sec
x = cpu time needed to execute a single
There appear to be a number of contradictory items
in your posting; presumably due to the passage of
time and the number of variations and experiments
that have take place.
You seem to indicate that a simple select on a single
table using an IN list takes 2 seconds to complete,
but the time
-L
Sent: Wednesday, April 17, 2002 11:48
PM
Subject: Re: HIGH CPU WITH MULTIPLE
CONCURRENT USERS (long)
Never mind. I just saw that oracle was able to
reproduce it internally.
- Original Message -
From:
Richard
Eastham
To: Multiple
). For us the degradation is so high
that we are not able to run 150 transactions per minute on a 4-processor
server. The simulation within oracle also supports this degradation
_
15-APR-02 22:09:08 GMT
Pasting information into the tar on bug:2321553 Abstract: HIGH CPU WITH
MULTIPLE
Message -
From:
Aponte, Tony
To: Multiple recipients of list ORACLE-L
Sent: Wednesday, April 17, 2002 11:48
PM
Subject: RE: HIGH CPU WITH MULTIPLE
CONCURRENT USERS (long)
I
think you are running into the stampeding herd phenomenon. I'm
suspicousof the low value
How many extents for the table?
- Original Message -
From:
Richard Eastham
To: Multiple recipients of list ORACLE-L
Sent: Tuesday, April 16, 2002 11:58
PM
Subject: HIGH CPU WITH MULTIPLE
CONCURRENT USERS (long)
A co-worker is having a fairly serious
Never mind. I just saw that oracle was able to
reproduce it internally.
- Original Message -
From:
Richard Eastham
To: Multiple recipients of list ORACLE-L
Sent: Tuesday, April 16, 2002 11:58
PM
Subject: HIGH CPU WITH MULTIPLE
CONCURRENT USERS (long
more thing, thanks for posting your problem with such clarity and supporting
detail.
Tony
Aponte
-Original Message-From: Richard Eastham
[mailto:[EMAIL PROTECTED]]Sent: Wednesday, April 17, 2002 1:58
AMTo: Multiple recipients of list ORACLE-LSubject: HIGH
CPU WITH MULTIPLE
the degradation is so high that we are not able to run 150
transactions per minute on a 4-processor server. The simulation within oracle
also supports this degradation
15-APR-02 22:09:08 GMTPasting information into the tar on
bug:2321553 Abstract: HIGH CPU WITH MULTIPLE CONCURRENT USERS since
We have problem with oracle process ora_snp0 that causes high CPU and core
files on our system. Does someone have idea how to fix it? Please help.
Below is information captured from a trace file:
Dump file /export/home/oracle/product/8.0.5/rdbms/log/ssdb_snp0_16657.trc
Oracle8 Enterprise
What job(s) are you running?
and if you strings core | head -40, what do you get?
Inquiring Minds want to know,
etc.
-Original Message-
Sent: Tuesday, January 29, 2002 12:06 PM
To: Multiple recipients of list ORACLE-L
We have problem with oracle process ora_snp0 that causes high
Hi,
I would like help with this problem I am grappling with since yesterday. One
of the test boxes with Sun Solaris 2.7 OS and Oracle 8.1.7.2 on it has
started showing 97% and above CPU usage mostly by Oracle . It has slowed
down the application to a point where the testers can not test. Does
Has there been any changes in the applications? I had someone put in new
code this weekend and it sent my CPU up over 80%. When I saw that Monday
morning, I had the developer take the code out and my CPU dropped back down
to around the 5 - 10%. From my little experience I have quickly learned
Uma - Can you identify which Oracle process is using the CPU?
-Original Message-
Sent: Tuesday, January 22, 2002 9:20 AM
To: Multiple recipients of list ORACLE-L
Hi,
I would like help with this problem I am grappling with since yesterday. One
of the test boxes with Sun Solaris 2.7 OS
Yeah there was a sudden burst of activity in data insert, updates and
deletes over the weekend as a result of a new app module being tested. But
the client app uses Java code and is not in the database. So its nothing I
can ask them to remove. On the other hand sudden bursts of data activity is
/
Jared
Uma Mohoni [EMAIL PROTECTED]
Sent by: [EMAIL PROTECTED]
01/22/02 07:20 AM
Please respond to ORACLE-L
To: Multiple recipients of list ORACLE-L [EMAIL PROTECTED]
cc:
Subject:High CPU usage by Oracle
Hi,
I would like help with this problem I am
Check if there are some cache buffer chains contention
through v$session_wait and v$latchname. Maybe some
sqls have changed their plans. Because of index
dropped or change of code.
Regards.
--- Uma Mohoni [EMAIL PROTECTED] wrote:
Yeah there was a sudden burst of activity in data
insert,
PROTECTED]
cc:
Subject:High CPU usage by Oracle
Hi,
I would like help with this problem I am grappling with since yesterday.
One
of the test boxes with Sun Solaris 2.7 OS and Oracle 8.1.7.2 on it has
started showing 97% and above CPU usage mostly by Oracle . It has
37 matches
Mail list logo