> Real-life example : using Siebel EIM, by placing EIM tables into separate > buffer pools, I saw a small advantage, say 5 - 10 % in buffer cache latch > reduction and more efficient use of cached IO. But after tuning the > structures so that I reduced logical IO's, I saw a 2000% throughput > improvement of EIM, to the amazement of all skeptics on the project (also > bumped up initrans and ran multiple parallel streams). So > prioritize where > you spend your tuning efforts. Reduction of logical IO = biggest bang for > buck ! > > Getting off my soapbox now. Lots to do. > Ciao : > > Ferenc Mantfeld
Ok, 5-10%, before or after everything else is addressed? I totally agree with where the biggest bang for the buck comes from and where focus should be spent. But let's assume one has reduced all of the needless work and poor algorithms, that the CBO is making great decisions in 99.999% of the cases, the physical side is nailed down perfectly, the data is optimally organized, etc. Just how much more could be squeezed out by making great use of the multiple buffer pools? We have all seen the cache effect -- the query that completes much faster when immediately run again (now was that the buffer cache or your EMC cache and how does all that play together, x$bh here we come ;-)). Just curious how much more could be gained, not that we are at that point yet ;-) Larry Elkins -- Please see the official ORACLE-L FAQ: http://www.orafaq.com -- Author: Larry Elkins INET: [EMAIL PROTECTED] 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: [EMAIL PROTECTED] (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).
