On Thu, 17 Aug 2006 01:28:34 -0400, Jim Mulder <[EMAIL PROTECTED]> wrote:
>>... > I looked at the PMR, and it does not mention a dump of the >address space in question after the CANCEL was done. A dump would be >required to figure out what is going on before trying to decide if there >is a problem which could be fixed. > > In general, the need for FORCE does not necessarily mean there is a bug. >It depends on the specifics of the situation in question. >... I didn't get notified in time to suggest a dump. The story I got was that this happened on 11 of 12 LPARs being brought down last weekend. On the 12th LPAR they tried cancelling the offending address space before starting the shutdown process. This worked. I agree that need for FORCE definitely depends on the situation, but I also think we identified a consistant situation where the FORCE is needed. It seems doubtful to me that the same non-bug situation arose on 11 LPARs. It's also possible that bug (if it is a bug) lies somewhere in MVS rather that in this one particular program, but no other address space had the problem. In the future we are going to do the cancel early in the shutdown procedure so we may not ever run into this again. And I may not convince anyone here that this problem is worth reproducing. In other words, I may never get the needed dump. Are you suggesting I should not have opened a PMR in this case? Pat O'Keefe ---------------------------------------------------------------------- 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
