Hi,

At z/OS 1.9 we had a problem with an OS checkpoint hang where the job
had to be FORCEd from the system.  We did a RESTORE of PE PTF UA38209 as
the immediate resolution.  We are tracking OA25786 as a permanent
resolution though we were told it was rare enough we were unlikely to
run into the same problem again. 

IHJ001I EMO40M,JS001.PS004 (CHKPNT,07E3,885161) INVALID CHECKPOINT
C0000210 (36-6-8) MODULE = IHJQCP32

Evidently our applications didn't both to ABEND if the checkpoint did
not succeed.

I reviewed the jobs that were taking checkpoints and recommended that
the OS checkpoint logic be removed as these jobs almost all of them in
the job series normally got INVALID CHECKPOINT and ran so quickly that
check points where not justified.  The application developers
subsequently did remove those checkpoint calls from the program and the
data sets from JCL.

My consideration is that OS checkpoints are less useful since any number
of services used makes them invalid as often does any involvement with
an allied address space that is a service provider.   We still make
appropriate use of IMS checkpoints and DB2 commits of course but have
for a while discouraged OS checkpoints. 

        Best Regards, 

                Sam Knutson, GEICO 
                System z Performance and Availability Management 
                mailto:[EMAIL PROTECTED] 
                (office)  301.986.3574              

"Think big, act bold, start simple, grow fast..." 

APAR Identifier ...... OA25786      Last Changed ........ 08/09/05
  FOLLOWING THE APPLY OF PTFS FOR OA22309, A HANG OCCURS AS A
  RESULT OF A WAIT IN IGG019BB + 300 DURING CHECKPOINT PROCESSING.
 
  Symptom ...... WS INCORROUT         Status ........... OPEN
  Severity ................... 2      Date Closed .........
  Component .......... 5695DF109      Duplicate of ........
  Reported Release ......... 190      Fixed Release ............
  Component Name CHECKPOINT REST      Special Notice    PE     HIPER
  Current Target Date ..08/09/25      Flags
  SCP ...................               FUNCTIONLOSS
  Platform ............                 PERVASIVE
 
  Status Detail: REVIEW - APAR solution is being reviewed.
 
  PE PTF List:    UA38208 UA38209 UA38207
 
  PTF List:
 
  Tentative Affected Releases and Current Relief Available:
  Release 1K0   : Relief is available in the form of: APARFIX
  Release 180   : Relief is available in the form of: APARFIX
  Release 190   : Relief is available in the form of: APARFIX
 
 
  Parent APAR:
  Child APAR list:
 
 
  ERROR DESCRIPTION:
    After applying the PTFs for APAR OA22309, jobs that issue a
  Checkpoint may result in a Wait in IGG019BB + x'300'.
    To verify this is the problem, check to see if a Checkpoint
  SVC (SVC63 / 0A3F) is outstanding on the RB chain.
  ADDITIONAL SYMPTOM
  ------------------
    It is possible for the Job which issued the checkpoint/restart
  to have an Exclusive ENQ on major SYSZJES2 and minor
  SJB.hexaddress.  If a TSO user attempts to view the
  checkpoint/restart job via SDSF, the TSO user will hang in a
  wait for a Shared ENQ.  The TSO user can be cancelled to get out
  of the ENQ wait.
 
 
  LOCAL FIX:
  Restore the applicable PTF or see Level 2 for the applicable
  APARFIX.


-----Original Message-----
From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED] On
Behalf Of BOB COSBY
Sent: Wednesday, September 10, 2008 1:55 PM
To: [email protected]
Subject: CheckPoint issue/question

Has anyone running z/OS Release 1.9 had a problem with batch jobs
utilizing Check Point abending?
If the error code coming from the ibm checkpoint is greater than 4 and
we have tried at least 10 times to take a checkpoint the job abends with
a U0500
IHJ000I CHECKPOINT N95P4400,N95P4400.STEP0020 (CHKPOINT) NOT TAKEN (14)
MODULE = IHJACP01


====================
This email/fax message is for the sole use of the intended
recipient(s) and may contain confidential and privileged information.
Any unauthorized review, use, disclosure or distribution of this
email/fax is prohibited. If you are not the intended recipient, please
destroy all paper and electronic copies of the original message.

----------------------------------------------------------------------
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

Reply via email to