Re: checkpoint spikes

"Arshad, Shehla" wrote: > > Mark, > > I have been making use of LRU Contention script that you provided > (Thanks!!), to check LRU contention on one of our servers. LC (spi*100/buf) > is mostly more than 5 at busier times. We have increased the number of LRUS > and CLEANERS to 127 but it seems there have been little or no effect on LRU > contention. What are we missing here? What else can be changed? The first thing to note is that having LRU contention doesn't necessarily mean you have a problem. If you have a problem, such as long checkpoints, then one of the things to look at is LRU contention. Tell us what your problem is and we'll tell you what to look at. > We are running IDS 7.30.FC7 on HP-UX 11 with 10 cpus given to this instance. > There are 300000 Buffers, 40 AIOVPS, 135 Chunks, 28 disks. Please advise. I would imagine that you can increase LRUS to 512 on that platform, CLEANERS I guess is at the limit, barring the annoying message at 128. ;-) I have found that you can usually expect SOME LRU contention with multiple CPU VPs, if that is what you mean by 10 CPUs. But as I said before, this is not necessarily a problem. If you do have a problem with checkpoint duration, then you might look at increasing the number of AIO VPs. I presume you have KAIO switched off because it's HP. 40 for 135 chunks doesn't sound right even though you only have 28 disks. Start at, say, two per disk and see if it improves things. If it does, then see how high you can push it. Cheers, -- Mark. +----------------------------------------------------------+-----------+ | Mark D. Stock mailto:mdstock@mydas.freeserve.co.uk |//////// /| | http://www.informix.com http://www.informixhandbook.com |///// / //| | http://www.iiug.org +-----------------------------------+//// / ///| | |What year 2000 bug? year 2000 bug? |/// / ////| | |year 2000 bug? year 2000 bug? year |// / /////| | |2000 bug? year 2000 bug? year 1900 |/ ////////| +----------------------+-----------------------------------+-----------+