OA64382: ZAA INTERIM FIX - Fixing several configuration and functional errors.

Obtain the fix for this APAR.

APAR status

• Closed as program error.

Error description

• ZOA INTERIM FIX - Service level 5.1.0.9a - Fixing several
  configuration and functional errors.
  

Local fix

Problem summary

• ****************************************************************
  * USERS AFFECTED: Z Anomaly Analytics with Watson 5.01 users   *
  ****************************************************************
  * PROBLEM DESCRIPTION: This PTF fixes some limitations         *
  *                      described in the Problem Summary.       *
  ****************************************************************
  * RECOMMENDATION: APPLY and activate this PTF.                 *
  ****************************************************************
  
  (1) The metric-based machine learning scorecard user interface
      does not distinguish between mutiple subsystems with the
      same subsystem ID for MVS systems within the same sysplex.
      The scorecard user interface only displays one of the
      subsystems.
  
  (2) Metric-based machine learning is unable to train a model for
      an MQ queue manager that is not part of a queue sharing
      group. In this case there is no data in the SMF115_2_CF
      table.
  
  (3) The database=cleanddata_history and database=cleandata_ml
      metric-based machine learning commands fail with message
      AIHW108E when the enterprise data warehouse is using Db2 for
      z/OS version 13 or some more recent versions of Db2 for z/OS
      version 12.
  
  (4) The define=find metric-based machine learning command only
      finds IMS subsystems that have transaction data (data in the
      IMS_56FA table).
  
  (5) The define=create metric-based machine learning command
      fails with a keyError for 'IZOA_DBLOC'.
  

Problem conclusion

• (1) The metric machine-learning scorecard user interface
      displays all subsystems, even those with the same subsystem
      ID for MVS systems within the same sysplex.
  
  (2) Metric-based machine learning is able to train a model for
      an MQ queue manager that is not part of a queue sharing
      group. The train job will still report a FAILED status
      because it is unable to train a model for the
      coupling_facility pipeline, but it trains models for the
      log, queue_manager, buffer_pool, and page_set pipelines. In
      this case if all expected models have been trained the
      FAILED status can be disregarded.
  
  (3) The database=cleanddata_history and database=cleandata_ml
      metric-based machine learning commands work correctly on all
      versions of Db2 for z/OS.
  
  (4) The define=find metric-based machine learning command is
      changed to look for subsystems in the IMS_4502 table rather
      than the IMS_56FA table. Any IMS subsystem that has region
      data is found.
  
  (5) The define=create metric-based machine learning command no
      longer fails with a keyError.
  

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:

  UJ92349

Modules/Macros

• AIHIMGPX
  

Fix information

• Fixed component name

  Z ANAL ANALYTIC

• Fixed component ID

  5698ABH06

Applicable component levels

• R510 PSY UJ92349

     UP23/03/10 P F303

Fix is available

• Select the PTF appropriate for your component level. You will be required to sign in. Distribution on physical media is not available in all countries.