Hi,Alan
Actually, there are no workload on both lpar.And interest thing is ,the fail one always is XP1K. ------------------ Original ------------------ From: "Staller, Allan"<[email protected]>; Date: Wed, Sep 12, 2012 09:10 PM To: "IBM-MAIN"<[email protected]>; Subject: Re: CTC connection based zos plex guest failed with IXC402D Sounds like a performance related issue. Is the failing LPAR(guest) one with a low relative performance rating compared to the partner? There are numerous documented instances of a poorly performing member of a SYSPLEX (due to low weight/VM dispatch priority) causing issues. Even happened within z/OS itself (see IEAOPT00 parameter ERV) which has do due with low priority tasks holding resources needed by higher priority tasks. HTH, <snip> I established a base plex beyond VM.It consists of two zos guests connected with each other by ESCON CTC. But...one of them fails every half day...what's a poor implemention! Every time it give a message like that when it died: IXC402D XT1K LAST OPERATIVE AT 22:23:06. REPLY DOWN AFTER SYSTEM RESET, OR INTERVAL=SSSSS TO SET A REPROMPT TIME. and the vm console got this: 00: HCPGSP2629I The virtual machine is placed in CP mode due to a SIGP stop from CPU 01. 00: HCPGIR450W CP entered; disabled wait PSW 00020000 80000000 00000000 000200A2 01: HCPPCM6533A Following is a priority message received by the service proces sor - use the VINPUT command to respond: 01: *IXC220W XCF IS UNABLE TO CONTINUE: WAIT STATE CODE: 0A2 REASON CODE: 01: 020, 01: THIS SYSTEM RECOGNIZED THAT IT WAS REMOVED FROM THE SYSPLEX </snip> ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
