I'm supposing that you have AUTOIPL coded in DIAGxx. AUTOIPL causes automatic IPL in a situation that would otherwise leave you in a wait state. We actually had this problem repeatedly early in R13 because the system would run totally out of real storage, a problem that was eventually fixed by APAR. The main point is that DIAGxx should also include the SADMP parameter as shown below. It requires that you configure standalone dump to work without manual intervention, a good idea in any case. You should get an SAD for most any wait state code other that V XCF,OFF.
AUTOIPL SADMP(hhhh,SMSYSC) MVS(LAST) . . . J.O.Skip Robinson Southern California Edison Company Electric Dragon Team Paddler SHARE MVS Program Co-Manager 626-302-7535 Office 323-715-0595 Mobile [email protected] -----Original Message----- From: IBM Mainframe Discussion List [mailto:[email protected]] On Behalf Of Vernooij, CP (ITOPT1) - KLM Sent: Monday, March 23, 2015 7:10 AM To: [email protected] Subject: Re: Unscheduled automatic reIPL on z10 Or if the automation tool does regularly re-IPL systems, it could have decided to do so at that moment for an unknown reason. Kees. -----Original Message----- From: IBM Mainframe Discussion List [mailto:[email protected]] On Behalf Of Lizette Koehler Sent: 23 March, 2015 15:03 To: [email protected] Subject: Re: Unscheduled automatic reIPL on z10 Did you seen any device contentions? Sometimes if a DASD has a reserves and it is a critical volume, like a Jes CKPT dataset, then the system only allows so many seconds to get at it. If it does not, the system could be considered non responsive. Any SVC Dumps during that time? How intact is your syslog? If syslog stopped minutes before the event, then you might not see the details. Check SMF Data as well for PDS/E, zFS or other statistics. If you have OPS/MVS or other automation tool, it might have captured messages you could review. Lizette > -----Original Message----- > From: IBM Mainframe Discussion List [mailto:[email protected]] > On Behalf Of Beesley, Paul > Sent: Monday, March 23, 2015 6:44 AM > To: [email protected] > Subject: Unscheduled automatic reIPL on z10 > > An hour or so ago, one of our z/OS systems decided it had enough of > life and re-IPLed itself. > > In the HMC hardware log we see this: > A logically partitioned mode event occurred. Select details for information. > Selecting Details gives: > A program directed re-IPL has been initiated for partition XXXX. > > First thought was that someone had done a V XCF,SYSNAME=XXXX,REIPL but > that would have generated a wait state record in the hardware log, And > in any case I have checked all of the other 13 systems in the sysplex, > and there is no evidence of this happening (and in fact, a CMDEXIT > would have prevented this from another system) > > There are no clues in this system’s syslog as to what triggered this. > Nothing in LOGREC either. > On the other systems we can see the usual messages indicating a system > has left (ISG378I, MIM saying a system may be inoperative, etc.) but > nothing else useful. > > Has anybody else experienced this? > This is a z/10 (2098-Z03) - I don’t know what level of microcode. All > 14 sysplex members being at z/OS 1.13, mostly RSU1308 or higher. > > > Regards and thanks > Paul ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
