RE: Please Advice on Performance Tuning

From: Hallas, John <HallasJ_at_logicae.com>
Date: Fri, 27 Apr 2001 03:03:39 -0700
Message-ID: <F001.002F3CF3.20010427011543@fatcity.com>

FOR YOUR INFORMATION ESIS and EPFAL are now part of Logica. The Internet email addresses of the staff has changed to the following - lastnameinitial_at_logica.com eg SMITHK_at_logica.com. Emails using the old format will continue to be delivered until 30th June 2001.

Naba,
Whilst you state that both schemas have the same initial and next etc you don't so whether they have the same number of extents. One thing I have seen a couple of times is a table that has a HWM due to rows being added and deleted. A FTS reads upto the HWM despite the fact there may be only a few rows in then table. Just a thought.
As David Barbour suggests, tracing is the best option to identify where the time is going. Perhaps an easier method of putting trace on each piece of code is to put the sql_trace = true parameter into your init.ora, restart the database and run the 30 minutes worth of code. Analyze the large trace file that will be produced (tkprof) and focus on the piece(s) of code with largest elapsed times .

HTH John

-----Original Message-----
Sent: 27 April 2001 06:25
To: Multiple recipients of list ORACLE-L

Hi all,

We have got a Payroll Application develeoped in-house.

It takes 30-34 Minutes to Run this Batch Application. General feeling is - it should be able to process it within 10-15 Minutes may be less than that.

I am to look into this problem and give a solution to it.

Steps I had followed as

Import the payroll user to another Oracle User in the same database.
Removed unnecessary Index , Put parallelism into few tables by Alter table tab1 Parallel(DEGREE 5)
Analyze the tables with compute statistics;
Run the payroll without checking any inefficient SQL or wrong programming logic in the those Packages( It has only two Package in it no other stand alone Procedure or function).

Payroll under this new Oracle User runs in 22-24 Minutes, 6-9 Minutes less. So, I asked Payroll Person to remove those unnecessary Index, asked him to put parallelism into those tables which I had done. He runs the Payroll with no Improvement at all. I asked again to drop those table and recreate it with new storage parameter same as the one created at New User. Still no visible Improvement.

At this stage every table in both user has same storage parameter,same index ,and also analyzed. The question is why in One User it runs in 22-24 Min and in another 30-34 Min ?

Now Please advice me what do I check or to do, so that the it time takes to run comes down to 22-24 Min, same as the new user.

Oracle 8.1.4
Optimize goal : choose
Biggest table haiving record less than 60000 rows.

Thanks in Advance

Naba

This e-mail and any files transmitted with it, are confidential to Logica and are intended solely for the use of the individual or entity to whom they are addressed.
--

Please see the official ORACLE-L FAQ: http://www.orafaq.com
--

Author: Hallas, John
INET: HallasJ_at_logicae.com

Fat City Network Services    -- (858) 538-5051  FAX: (858) 538-5051
San Diego, California        -- Public Internet access / Mailing Lists
--------------------------------------------------------------------

To REMOVE yourself from this mailing list, send an E-Mail message to: ListGuru_at_fatcity.com (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB ORACLE-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing). Received on Fri Apr 27 2001 - 05:03:39 CDT