Re: concurrency problem in 5.07

In article <364A546F.54AB198@hotmail.com>, June Tong <june_t@hotmail.com> writes >robert schmidt wrote: > >> when more than two queries run concurrently, the response time is >> highly degraded. All of the data that the queries are reading >> is in informix buffer cache, so there is no i/o time. >> The inquiries are read only (cursor stability), >> so not too much locking should be taking place.... > >I'd be interested to see what kind of locking IS taking place. That, and the >latching. Have you looked at an onstat -sk? > > >> So far as I can tell, there is nothing interesting in 'sar' reports >> when the system is showing this degraded performance. The overall >> CPU utilization actually goes down as more processes are run >> concurrently. > >Then it would appear they are all waiting for something. Actually, let me add >an onstat -u to the list of interesting output. > onstat -p 10 seconds apart would be interesting, see what is changing on the server and which counts are going up... PS How many physical CPUs do you have, how many disks, sar -u 2 2 and sar -d 2 2 and sar -p 2 2 show what? Maybe you are paging/swapping.. > >> The only idea we have at this time is adjusting the 'SPINCNT' configuration >> parameter... Perhaps double it to 600. Any thoughts on this idea? > >That would imply that they are waiting for a lock or a latch. First, you >should see what they are waiting for. > >June >-- >june_t@hotmail.com >Grounded in Palo Alto, living on Oreo's > > -- David Williams Maintainer of the Informix FAQ Primary site (Beta Version) http://www.smooth1.demon.co.uk Official site http://www.iiug.org/techinfo/faq/faq_top.html I see you standin', Standin' on your own, It's such a lonely place for you, For you to be If you need a shoulder, Or if you need a friend, I'll be here standing, Until the bitter end... So don't chastise me Or think I, I mean you harm... All I ever wanted Was for you To know that I care