Very interesting. In several shops I have just dealt with the message in a timely fashion. We indeed do the v xcf,sysname,offline and every time one of the other systems gets the " error " message. It does seem like a kludge to me and I've suggested we set up ZAKK to automagically issue the "down". Note now tho that we are getting IXC102A and your note references IXC402D. Furthermore, we don't get a 90 second grace period. We get the error message immediately and in the 60 seconds it takes to reply down we get the 'really ugly problems". On startup the lpar happily rejoins the (mono?)plex with no problems. If you solve this, Barbara, I will bake you some cookies. One other poster suggested the distributed systems connected to the other lpar in this plex may be confused due to VIPA definitions, but I have not tracked that down yet.
1) SHUTSYST a) PF9 = K E,1 2) Stop or cancel tasks that refuse to recede as needed. 3) Use Force as a last resort only. a) ALL AVAILABLE FUNCTIONS COMPLETE 4) $PJES2 a) JES2 ENDED 5) Z EOD 6) V XCF,SYSTS1,OFFLINE a) A confirmation message will prompt you b) R__,sysname=systs1 7) Wait for the screen to blank and a few messages will reappear. a) TYPE 2096 - S07 Mfg = IBM ... 8) On HMC a) Click RESET b) Click OK 9) On one of the other systems that are still up (every second counts!) a) nn IXC102A XCF IS WAITING FOR SYSTEM SYSTS1 DEACTIVATION. REPLY DOWN WHEN MVS ON SYSTS1 HAS BEEN SYSTEM RESET b) R__,DOWN Ken Klein Sr. Systems Programmer Kentucky Farm Bureau Insurance - Louisville [email protected] 502-495-5000 x7011 -----Original Message----- From: IBM Mainframe Discussion List [mailto:[email protected]] On Behalf Of Barbara Nitz Sent: Wednesday, July 08, 2009 3:03 AM To: [email protected] Subject: Re: Sysplex timeout problems. You're NOT supposed to get any message to which you have to reply DOWN! The correct way to shut down a system in a sysplex is to ALWAYS ALWAYS use vary xcf,sysname,offline. Then reply with the name of that system again. (This is true for monoplexes, also!) Remember that the IXC402D (reply down when mvs has been system reset) is an ERROR condition message. Issued by XCF on another system after 90 seconds when the system in question does not update its couple CDS heartbeat anymore (and does not communicate via XCF signalling anymore). The reason for that failure to communicate *should* always be a 'real' error on that system and never due to that system just getting its icon dragged in the course of normal shutdown. The purpose of vary xcf offline is to tell all connectors to all XCF groups in that sysplex that a system will shortly be going away and do cleanup on behalf of that system. Failure in that cleanup can lead to really ugly problems during restart. regards, Barbara Nitz ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
