Performing fallback

If a problem occurs after HCL Workload Automation for Z has been active as a production system for some time, and the problem is serious enough, you might need to stop the new system and return the workload to the previous system. You can do this by using a procedure called fallback, if the HCL Workload Automation for Z data sets are usable.
Note: If on the primary controller the Step Awareness feature was active and you want to keep it in the previous system, ensure that you convert the EQQSTDS data set.
The fallback procedure is as follows:
  1. Run the EQQPCS01 and EQQPCS02 jobs to allocate new data sets for the old production system. The current data sets, or a copy, used by the HCL Workload Automation for Z systems should be kept for problem determination purposes.
  2. If required, close down the systems in the same way as during migration. This is required if the current plan on the OPCA system is intact and job tracking is working normally.
  3. If possible, create up-to-date data sets for the long-term plan and new current plan for the OPCA system.

    Do not submit a REPLAN job prior to shutdown, unless the PERMANENT option was used for SSCMNAME on the converted system, or if SSCMNAME was not specified.

    If SSCMNAME(EQQSSCM1,TEMPORARY) was used, message EQQX145E will be issued if a REPLAN job is started after the controller is shut down.

  4. Build VSAM data sets for the old system by running the EQQICNVS job to convert HCL Workload Automation for Z files to their previous format.
    Note: Before running EQQICNVS job, check that a daily plan batch process was not submitted on the system before shutting the controller down. Then, in EQQICNVS set the HCL Workload Automation for Z CP1, CX, and XD1 as the input, and NCP, NCX, and NXD as the output.
  5. Ensure that in the JTOPTS initialization statement you set CURRPLAN(NEW).
    Note: You might find useful to specify JOBSUBMIT(NO) in JTOPTS, so that work is not submitted when you start OPCA. After checking that the old system has started without errors, you can activate job submission by using the Service Functions dialog.
  6. Start the OPCA and OPCB systems again using the converted files and start OPCS.
  7. If the new current plan (NCP) data set is not fully up-to-date because you could not run the daily plan program, use the MCP dialog to update the status of operations to make the current plan up-to-date.
  8. Start the OPCC and OPCD systems again. Use the SSCMNAME parameter on the JTOPTS initialization to load the current subsystem communication module for the release to which you are falling back.