-------- Original-Nachricht -------- > Datum: Thu, 27 Dec 2007 10:44:38 +0100 > Von: Rob van der Heij <[EMAIL PROTECTED]> > An: IBMVM@LISTSERV.UARK.EDU > Betreff: Re: Paging sub system seems extremely slow
> On Dec 27, 2007 9:25 AM, Klaus Johansen <[EMAIL PROTECTED]> wrote: > > > Our VM system pages extremely slowly: a long inactive Linux guest is > paged in from DASD by z/VM with approximately 1 MB/sec in an otherwise almost > idle system. The overall system page rate is 400-800 pages/second – half > for reading pages and half for writing pages. We have 12 paging disks > (3390-9) distributed over 2 LCUs and attached with 6 FICON channels – neither > disks nor channels are in any way a bottleneck according to our performance > readings. > > I'm not familiar with your performance monitor. If yours includes > tuning support, then those folks will be able to help you. > But if you want to capture some raw monitor data, I'd be happy to > browse that and see whether I can explain what you see. Drop me a note > off-list if you're interested. Thank you, will do... > > I do spot a number there that might suggest your paging blocking is > too low (happens when your paging space is very full). I also see you > have MDC in XSTORE which is something you probably should not do > anymore. You haven't told yet what level of z/VM you run, and there's > a zillion of other questions I would have to ask when looking at this > (all answered by the raw monitor data). > Paging blocking does seem very low but our page volume utilization is maximum 23%. z/VM seems constantly to take a few pages here and there from all the guests all - maybe that's why the paging volumes become "fragmented". I have disabled MDC in XSTORE - doesn't seem to have any effect... We're running 5.2 service level 0701... > > At the same time the main storage is badly utilized: Performance Toolkit > reports a storage utilization of 65-75% and it reports strange values for > "Total real Storage" and "Total available" storage: namely 0kb. > > The "storage utilization" number does not have much value these days. > I suspect the low value might be due to MDC in main memory (which may > drive your paging up, but does not explain poor paging). > While I don't know whether the "total" comes from, it could very well > be that your MONDCSS is not configured right and causes you to miss > configuration data (I'm told even Performance Toolkit complains about > it when that happens). It's the "config sample size" that is set on > the MONITOR START command. You were quite right: The config sample size was to low, and Performance Toolkit did complain about it. The mentioned values and a few other now show up correctly. > > Rob > -- > Rob van der Heij > Velocity Software, Inc > http://velocitysoftware.com/ -- Psssst! Schon vom neuen GMX MultiMessenger gehört? Der kann`s mit allen: http://www.gmx.net/de/go/multimessenger?did=10