RE: performance questions

Sai - I think the whole point is to open this up for discussion / negotiation. My suggestion would be to agree with the business users on a "typical" query and the response time they expect. Ideally response time is measured at the user's terminal. But if necessary just for the database, you might get agreement on a typical query and the response time.

    As far as adding more stuff to the server, that is the whole point. If you can add more stuff and the SLA is maintained, great! Everybody's happy. But if you add users/instances/etc., and the SLA suffers, now you have a discussion point for the users. Either somebody can buy another server, or somebody can agree to a higher SLA, etc. The point is that you're talking, getting issues aired, rather than you guys saying @#$% users and the users saying @#$% DBAs.

Dennis Williams
DBA, 80%OCP, 100% DBA
Lifetouch, Inc.
dwilliams_at_lifetouch.com

-----Original Message-----
Sent: Tuesday, June 03, 2003 12:05 AM
To: Multiple recipients of list ORACLE-L

hi gurus  

this is a kind of query i have faced a few times in the recent past and which has really forced me to start this thread.  

as everyone knows, there is always what we call a SLA or in other words a service level agreement (may be called differently in different places) which infact means defining a time for any transaction to go thru in the database. This is very important in emvironments which handle transactions affecting sales or just normal queries against huge databases which helps a sales force or a front office customer support force.. Defining this is always a difficult task and i believe will keep changing as time goes on - factors like number of records,the number of databases running on a box(probably SLA was defined initially on a single box-single db kind of env and now the same box has more databases),memory,network,disk performance,number of transactions or can i say the load profile et al. there have been cases where i have been asked questions like why this query took more time than SLA when it was running ok sometime back. i find it very difficult to convince saying that ther! e are factors affecting this and not just explain plan et al(correct me if i am wrong) or in other words a scenario that says my test environment is running faster than prod
(everything on the db side are the same except the way the disks are
configured or the load profile on both dbs).  

here is my question? is there a way to determine this SLA. since it keeps changing how do we really determine it. there is a soltuion that comes right out saying abenchmark can help u do this but how do we extrapolate or assume that there was no benchmark done at the beginning how do we validate/dtermine this magic number.
i have some ideas on this but nothing is very concrete.  

can someone give me some feedback on this..if u feel that this is not a right question to be put in this forum i apologize but i would like to take this up with someone who is interested and i wouldnt use this mailing list for the same.  

thanks for ur time
sai

-- 
Please see the official ORACLE-L FAQ: http://www.orafaq.net
-- 
Author: DENNIS WILLIAMS
  INET: DWILLIAMS_at_LIFETOUCH.COM

Fat City Network Services    -- 858-538-5051 http://www.fatcity.com
San Diego, California        -- Mailing list and web hosting services
---------------------------------------------------------------------
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).