Troubleshooting Online Structure Change utility problems

If you receive error messages, you must identify the cause of the error, correct it, and, if necessary, restart the OSC job.

Before you start troubleshooting an OSC problem, ensure that your environment is correctly set up for the Online Structure Change utility. If any of the following conditions is not satisfied, the OSC driver job ends with errors:

  • Make sure that, for each ADS, the number of copies is less than seven. The Online Structure Change utility registers the shadow ADS as a copy during the job, and MADS (Multiple Area Data Sets) can only maintain up to seven copies of each area.
  • Make sure that all IMS systems in the same IMSplex where you run OSC have the executable load library of IMS HP Fast Path Utilities (SHFPLMD0) in the STEPLIB concatenation.
  • Make sure that the IMS DBCTL system is started with the new ACB definition.
  • Define a new OSC internal job data set if it does not exist. If it already exists, make sure that it has enough space. Also make sure that this data set is specified on the JCLDSN keyword and that the data set contains the OSC internal jobs that were generated and submitted during the OSC process.
  • Define a new ENQ data set if it does not exist. If it already exists, make sure that it has enough space. Also make sure that this data set is specified on the ENQDSN keyword and that the data set contains the members that were generated during the OSC process.
  • Define a new OSC internal job log data set if it does not exist. If it already exists, make sure that it has enough space. Also make sure that the prefix of this data set is specified on the LOGDSN keyword and that the data set contains the OSC internal job logs collected during the OSC process.
  • If ORIGADS_PROC=RENAME is specified, delete or rename the VSAM data set whose name is "original ADS + .T".

For details, see Running the Online Structure Change utility.

The following sections explain how to troubleshoot and restart Online Structure Change utility jobs.

When message HFPT0512A is issued

When the ERRORACT=WTOR option is enabled and one of the OSC internal job fails, the utility issues WTOR message HFPT0512A. When this message is displayed, follow these steps to restart the OSC driver job:
  1. From the HFPT0512A message, identify the OSC internal job that failed.
  2. Locate error messages in the corresponding job log and identify the cause of the error.
  3. Correct the error. If the error occurred in one of the following OSC internal jobs, follow the instructions and correct the failed OSC internal job in the internal job data set (specified by the JCLDSN keyword).
    If xxxxxxC0 (COPYSADS) failed
    1. Check the status of the shadow ADS in the production DBRC.
      • If the shadow ADS is in AVAIL status, separate the shadow ADS from the production IMS online system by issuing the /STOP ADS command, then unregister the shadow ADS from the production DBRC.
      • If the shadow ADS is in UNAVAIL status, unregister the shadow ADS from the production DBRC.
    2. Submit the xxxxxxF0 (DEFSADS) internal job to define the shadow ADS.
    3. Submit the xxxxxxI0 (INITSADS) internal job to initialize the shadow ADS.
    If xxxxxxxR0 (RELOAD) failed
    Check the restructured ADS and UIC in the DBCTL DBRC. If the restructured ADS and UIC are registered in the DBCTL DBRC, submit the xxxxxxB0 (UNDDBCTL) internal job.
    If xxxxxxM0 (OLC) failed
    Issue the TERMINALTE OLC command to terminate the global online change.
    If xxxxxxN0 (RENAME) failed
    Check whether the original ADSs exist.
    • If ORIGADS_PROC=DELETE and the original ADSs do not exist, remove the DELETE ORIGINAL ADS job step from the xxxxxxN0 internal job.
    • If ORIGADS_PROC=RENAME and the original ADSs do not exist, remove the RENAME ORIGINAL ADS job step from the xxxxxxN0 internal job.
  4. Rerun the failed OSC internal job and check the corresponding job log.
    • If the error is resolved, reply to the HFPT0512A message with C. The OSC driver job restarts.
    • If the OSC internal job fails, repeat steps 3 and 4.

OSC driver job ended abnormally when RUNTYPE=GENJCL

If the OSC driver job was submitted with the RUNTYPE=GENJCL option and the job failed, follow these steps:
  1. Review the OSC driver job log to obtain information about error.
  2. Correct the error conditions.
  3. Submit the OSC driver job.

OSC driver job ended abnormally before the takeover phase

Follow these steps if the OSC driver job was submitted with the RUNTYPE=FULL or RUNTYPE=RUNJCL option and ended abnormally before the takeover phase. The following steps are valid if the xxxxxxD0 (DBRPROD) internal job was not submitted.
  1. Submit the xxxxxxG0 (SAVELOG) job, which is generated in the internal job data set that the JCLDSN keyword specifies, to collect the job logs of the OSC driver job and OSC internal jobs.
  2. Check the job logs in the job log data set that the LOGDSN keyword specifies to obtain information about the error.
  3. Correct the error conditions.
  4. Check the status of the shadow ADS in the production DBRC.
    • If the shadow ADS is in AVAIL status, separate the shadow ADS from the production IMS online system by issuing the /STOP ADS command, then unregister the shadow ADS from the production DBRC.
    • If the shadow ADS is in UNAVAIL status, unregister the shadow ADS from the production DBRC.
  5. Check the restructured ADS and UIC in the DBCTL DBRC. If the restructured ADS and UIC are registered in the DBCTL DBRC, unregister them.
  6. Issue the /START PGM ALL command to the DBCTL region.
  7. Submit the OSC driver job.

OSC driver job ended abnormally during the takeover phase

Follow these steps if the OSC driver job was submitted with the RUNTYPE=FULL or RUNTYPE=RUNJCL option and ended abnormally during the takeover phase. The following steps are valid if the xxxxxxD0 (DBRPROD) internal job was submitted.

  1. Submit the xxxxxxG0 (SAVELOG) job, which is generated in the internal job data set that the JCLDSN keyword specifies, to collect the job logs of the OSC driver job and OSC internal jobs.
  2. Check the job logs in the job log data set that the LOGDSN keyword specifies to obtain information about the error.
  3. Correct the error conditions.
  4. Submit the following OSC internal jobs.
    • If xxxxxxM0 (issues the INITIATE OLC command) or xxxxxxX0 (issues the IMPORT command) has ended normally, submit the following OSC internal jobs:
      • xxxxxxO0 if this internal job is generated and was not submitted during the previous run
      • xxxxxxW0 if this internal job is generated and was not submitted during the previous run
      • xxxxxxN0 if this internal job is generated and was not submitted during the previous run
      • xxxxxxT0 if this internal job was not submitted during the previous run
      • xxxxxxL0 if this internal job was not submitted during the previous run
    • If xxxxxxM0 (issues the INITIATE OLC command) or xxxxxX0 (issues the IMPORT command) was not submitted or ended with an error, complete the following steps:
      1. If xxxxxxM0 ended with an error during or after the INITIATE OLC PHASE(PREPARE) command was issued and before the INITIATE OLC PHASE(COMMIT) command was issued, issue the TERMINATE OLC command to abort IMSplex-wide global online change that was initiated by the INITIATE OLC PHASE(PREPARE) command.
      2. If xxxxxxY0 ended successfully, restore the old randomizing routine or the old segment edit/compression routines to the IMSRESLB data set.
      3. Submit the xxxxxxV0 (STAORIG) job to bring the input DEDB and areas back online.
      4. If xxxxxxA0 ended abnormally, issue the /START PGM ALL command to the DBCTL region.
Start of change

OSC internal job xxxxxxJ0 (STAAPPLY) job ended with return code 20

If the DMACPRT step in the xxxxxxJ0 job ends with return code 20, follow these steps:

  1. Locate IMS Library Integrity Utilities error messages (FABL messages) for the DBRC region in the IMS DBCTL system.
  2. Compare database definitions (ACBs, randomizers, and segment edit/compression routines) used in the IMS DBCTL system and the new database definitions specified for the OSC driver job.
  3. If the ERRORACT=WTOR option is enabled, reply to the HFPT0512A message by typing A to end the OSC driver job.
  4. Update the database definitions to resolve ACB inconsistencies. If you update the ACB that is used in the IMS DBCTL system, restart the IMS DBCTL system.
  5. Rerun the OSC driver job.
End of change