tuning question

In looking at some performance tuning notes I noticed something that said "Increment the onconfig variable LRUS by 2 until the onstat -R percentage is below 60%." After examining the output and documentation for onstat -R I can't figure out what they were talking about and why you would want to do this. If you can shed some light on the subject I would greatly appreciate it!

In article <81ul38$rve$1@nnrp03.primenet.com>, "Top Cat" <cstefanick@ocsmgmt.com> wrote: > In looking at some performance tuning notes I noticed something that said > "Increment the onconfig variable LRUS by 2 until the onstat -R percentage is > below 60%." After examining the output and documentation for onstat -R I > can't figure out what they were talking about and why you would want to do > this. If you can shed some light on the subject I would greatly appreciate > it! Bogus advice. Where did you see that? The ratio of clean to dirty buffers in the onstat -R report is dependent ONLY on MIN/MAX_DIRTY and whether there are enough IO VPs and CLEANERS to keep up with the work. Check my tuning article in Informix Tech Notes Issue 8, Number 3 or a more recent article in Tech Notes (Tech Notes is available online on the Informix WEB Site) or get the excellent book on the subject by Joe Lumbley (Informix DBA's Survival Guide, 2nd Edition) or the soon to be available book by Ron Flannery (Informix Handbook), which includes my article (shameless plug I know). Art S. Kagel Sent via Deja.com http://www.deja.com/ Before you buy.