PM03351: EXCESSIVE CPU RUNNING ORF

APAR status

• Closed as program error.

Error description

• Excessive CPU running ORF - (more details to follow)
  

Local fix

Problem summary

• ****************************************************************
  * USERS AFFECTED: Users of IMS Online Reorg Facility V1.2      *
  ****************************************************************
  * PROBLEM DESCRIPTION: When running the BMP pause feature,     *
  *                      high CPU in GRS is detected.            *
  ****************************************************************
  * RECOMMENDATION:                                              *
  ****************************************************************
  This APAR is correcting the following problems:
  1.When running the BMP pause feature, high CPU in GRS is
  detected. The BMP pause detection gets invoked when the
  application does a CHKP call. At that time it checks if
  a pause request has been set by the ORF job.  This is done
  using GQSCAN. However, if the checkpoint frequency is very
  high, this results in heavy GQSCAN activity.
  The time interval of 10 seconds was not working properly,
  resulting in heavy CPU.
  2.No good indication which BMP pauser is active.
  The ORF BMP pause routine did not provide any messages.
  3.When running TAKEOVER(DELAY) and restarting later with
  TAKEOVER(YES) the DBDs are not restarted in the online system
  When the second TAKEOVER runs, the databases are currently
  stopped. This is then considered being the initial status and
  the databases are not restarted.
  

Problem conclusion

• The following changes have been made to correct the above
  mentioned problems:
  1.Code has been changed to correct the time interval.
  An optional DD-stmt has been provided to trigger the GRS
  activity display. If //$$GRS  DD  DUMMY is provided in
  the BMP JCL, a WTO is created each time the GQSCAN routine
  is entered.
  2.New messages have been introduced to show the start and the
  end of the application
  3.Code has been changed to recognize that this is a restarted
  takeover. The database status at that time is ignored. That
  will then restart the databases.
  

Temporary fix

Comments

APAR Information

• APAR is sysrouted FROM one or more of the following:

• APAR is sysrouted TO one or more of the following:

  UK53588

Modules/Macros

• HRFOLEV0 HRFOREO0 HRFOREO2 HRFOREO8 HRFRRCXX
  HRFRRC10 HRFRRC11 HRFRRC81 HRFRRC91 HRFXDYTR HRFYCCDF HRFYCC01
  HRFYCC02 HRFYCINI HRFYOF10 HRFYOF11 HRFYOF81 HRFYOF91 HRFYSTRT
  

Fix information

• Fixed component name

  IMS ONLINE REOR

• Fixed component ID

  5655H9700

Applicable component levels

• R120 PSY UK53588

     UP10/01/19 P F001