Subject: RE: CICS error
From: Randy Evans (revans@viaserv.com)
Date: Fri Apr 04 2008 - 12:29:48 EST
No apology is warranted - the existence of this problem and your mention of only a single CICS is how I (erroneously) arrived there. Since you *do* have a production CICS and this problem still occurs, then I suggest the problem lies more with the change management process. As you pointed out, the programmer "should" NEWCOPY, but it certainly appears that the job done isn't getting done in all cases. I think if you search back through these archives, you might find some suggestions from others on how to automate the NEWCOPY function in the same JCL that performs the LIBR COPY. If that isn't successful, you might pose a new query back on this list as to how others have skinned that cat lately. In my past experience, the process of moving programs test-to-production was made the responsibility of computer operations, using JCL provided by the systems programmers. The jobs involved included steps to perform the NEWCOPY using a home-grown utility. This process significantly improved CICS uptime statistics for production CICS but as you can guess, I don't recall anyone in the applications development group being much in favor of this change at the time (mid 80s). Randy Evans, Viaserv, Inc. > -----Original Message----- > From: owner-vse-l@Lehigh.EDU [mailto:owner-vse-l@Lehigh.EDU] On Behalf Of > Dan Gofton > Sent: Friday, April 04, 2008 11:05 AM > To: VSE Discussion List > Subject: RE: CICS error > > Don't know how I gave that impression. Sorry. > > We run separate LPAR's (and thusly, separate CICS') for development and > production. There are no shared, updatable disks between the two. The > development LPAR disks are defined as READONLY in production LPAR to allow > us to copy. When a program is ready for production, the programmer > executes > a LIBR COPY to copy from the development library to the production > library, > then "should" NEWCOPY. > > Dan Gofton > Systems Programmer > County of Saginaw, MI > 989 790 5525 > > > -----Original Message----- > > From: owner-vse-l@Lehigh.EDU [mailto:owner-vse-l@Lehigh.EDU] > > On Behalf Of Randy Evans > > Sent: Friday, April 04, 2008 12:41 PM > > To: VSE Discussion List > > Subject: RE: CICS error > > > > It sounds like you run everything (test/development and production > > users) in a single CICS. If so, I suggest you (re)consider > > implementing a separate CICS for production usage vs. > > test/development work, and implement change management > > procedures to restrict when/how/who can implement changes > > (new/updated application programs) in production. This won't > > change the frequency or severity of these types of problems > > in the test CICS, but should virtually eliminate these > > problems in production CICS if your change management > > procedures are adequate. > > > > > > Randy Evans, Viaserv, Inc. > > > > > > > > > -----Original Message----- > > > From: owner-vse-l@Lehigh.EDU > > [mailto:owner-vse-l@Lehigh.EDU] On Behalf > > Of > > > Dan Gofton > > > Sent: Friday, April 04, 2008 9:36 AM > > > To: VSE Discussion List > > > Subject: RE: CICS error > > > > > > And I buy this explanation for the majority of these messages, but I > > know > > > I > > > have had a least one time where there were many transactions having > > this > > > problem. The solution was to bounce CICS. > > > > > > Possibly some of the changes we've made over the last few > > months have > > > exasperated the issue: going from a page data set to NOPDS, only > > cycling > > > (normal) CICS once a week instead of every night, etc. > > > > > > Thanks for your help. > > > > > > Dan Gofton > > > Systems Programmer > > > County of Saginaw, MI > > > 989 790 5525 > > > > > > > -----Original Message----- > > > > From: owner-vse-l@Lehigh.EDU [mailto:owner-vse-l@Lehigh.EDU] On > > > > Behalf Of Kevin Corkery > > > > Sent: Friday, April 04, 2008 11:13 AM > > > > To: VSE Discussion List > > > > Subject: RE: CICS error > > > > > > > > Your solution is the answer. A program has had it's physical > > > > position in the library moved since the CICS startup. > > Any program > > > > that's moved into (or out of for that matter) of a library in the > > > > CICS search chain needs to have a NEW COPY (or PHASEIN) command > > > > issued against it. > > > > > > > > -----Original Message----- > > > > From: owner-vse-l@Lehigh.EDU [mailto:owner-vse-l@Lehigh.EDU] On > > > > Behalf Of Dan Gofton > > > > Sent: Friday, April 04, 2008 11:03 AM > > > > To: VSE Discussion List > > > > Subject: CICS error > > > > > > > > Periodically we have been getting the following errors in our > > > > Production > > > > CICS: > > > > > > > > C1 0088 DFHLD0203 SAGPROD LOADER SVC LOAD REQUEST FAILED > > DUE TO I/O > > > > ERRORS > > > > > > > > PROCESSING THE LIBDEF SEARCH CHAIN. > > 04/04/08 > > > > 10.56.59 > > > > C1 0088 DFHLD0002 SAGPROD A SEVERE ERROR (CODE X'1925') > > HAS OCCURRED > > > > IN MODULE > > > > DFHLDLD1. > > 04/04/08 > > > > 10.56.59 > > > > C1 0088 DFHME0116 SAGPROD > > 04/04/08 > > > > 10.56.59 > > > > (MODULE:DFHMEME) CICS SYMPTOM STRING FOR MESSAGE DFHLD0002 > > IS > > > > 10.56.59 > > > > PIDS/564805400 LVLS/411 MS/DFHLD0002 RIDS/DFHLDLD1 > > PTFS/VSE411 > > > > 10.56.59 > > > > PRCS/00001925. > > 04/04/08 > > > > 10.56.59 > > > > C1 0088 DFHDU0201 SAGPROD ABOUT TO TAKE SDUMP. DUMPCODE: > > > > LD0002 , DUMPID: > > > > > > > > 1/0018 > > 04/04/08 > > > > 10.56.59 > > > > C1 0088 0S24I AN SDUMP OR SDUMPX MACRO WAS ISSUED > > 04/04/08 > > > > 10.56.59 > > > > C1 0088 0S29I DUMP STARTED > > 04/04/08 > > > > 10.56.59 > > > > C1 0088 0S30I DUMP STARTED. MEMBER=DC100117.DUMP IN > > SUBLIB=SYSDUMP.DYN > > > > 10.56.59 > > > > C1 0088 1I51I DUMP COMPLETE > > 04/04/08 > > > > 10.57.17 > > > > C1 0088 DFHDU0202 SAGPROD SDUMPX COMPLETE. SDUMPX RETURN > > CODE X'00' > > > > 10.57.17 > > > > > > > > Sometimes we can narrow it down to an offending program, issue a > > > > NEWCOPY and the problem goes away. Other times it > > becomes necessary > > > > to recycle the CICS partition to alleviate the problem. Before I > > > > contact IBM, would anyone have any idea what may cause > > this problem? > > > > > > > > Dan Gofton > > > > Systems Programmer > > > > County of Saginaw, MI > > > > 989 790 5525 > > > > > > > > > > > > > > > > > > > > > > > > > >
This archive was generated by hypermail 2b25 : Fri May 16 2008 - 11:35:06 EDT