This is a z/VM 5.2 SLU 0501 system that had been up for 7.5 months (last IPL was 3/26/2006). I have not (yet) openned a PMR with IBM. As stated, the dump may be of little value since I had taken down the servers and th e problem wasn't occuring at the time. I'm still shaking my head that I
didn't bring them back up and *then* take the dump. So, the trace table has almost certainly wrapped and has nothing useful. Beyond that would probably require being able to identify some corruption in CP control blocks, which IBM may or may not be able to do. I did open a problem with the software vendor since the problem went away and came back in concert with their servers being taken down and coming back up. This software has been running fine (last updated 9/11/2006) an d I don't think that it caused the problem. I would like their confirmatio n that I didn't overlook something important in the performance reports tha t would point out an area to investigate different from what I found. Any PMR with IBM will wait until I hear back from the software vendor. Brian Nielsen On Wed, 15 Nov 2006 22:00:49 -0500, David Kreuter <[EMAIL PROTECTED] resources.com> wrote: >whoa - do you have a PMR open, or an APAR to track? is this zvm520? >David > > >-----Original Message----- >From: The IBM z/VM Operating System on behalf of Brian Nielsen >Sent: Wed 11/15/2006 5:11 PM >To: [email protected] >Subject: Re: [IBMVM] IPL Strangeness > >And I just did a PSW restart yesterday because I needed to IPL to fix a = > >severe performance problem and wanted a dump of the system. > >For some reason CP was spending enormous amounts of time doing instructi o= >n >simulation that seemed to be triggered by the virtual machines involved i= >n >collecting the monitor data. I don't yet know definitively what >happened. The T/V ratio for several CMS machines skyrocketed, but not s o= > >for the Linux guests. > >Unfortunately, I took the dump after terminating the servers in question = > >so it's probably of little use. > >Brian Nielsen > >On Wed, 15 Nov 2006 16:08:09 -0500, David Kreuter <[EMAIL PROTECTED] >resources.com> wrote: > >>PSW restart (system restart function from the HMC) indeed still lives o n= > >and works brilliantly. Just did one last week - for fun and practice. >Still results in a SVC002 abend code. Then read the dump with VMDUMPTL = > >after DUMPLOAD processing. >> >>David >>>-----Original Message----- >>>From: The IBM z/VM Operating System on behalf of George Haddad >>>Sent: Wed 11/15/2006 4:01 PM >>>To: [email protected] >>>Subject: Re: [IBMVM] IPL Strangeness >>> >>>If it does re-occur, maybe you can do a PSW-restart to get a dump >>>(assuming that's still an option in the newer VMs) >> >>======================== = >======================== == >======================= >======================== ========================= =======================
