PH62138: SPECIFYING A FIXED SIZE OR THE VALUE "AUTOMATIC" FOR PARAMETER "TEMP_WORKING_SPACE" IN THE JSON CONFIG FILE COULD BE IGNORED

APAR status

• Closed as documentation error.

Error description

• When specifying the setting "automatic" or a fixed value for
  parameter "temp_working_space" in the json configuration file,
  the expectation is to
  - get the specified disk space being taken constantly
  
   (pre-allocated on the file system (data pool)).
  - see a fixed value for TEMPSPACE1 as "in use" in the
  
   SMF data.
  
  After upgrading to maintenance level 7.5.11.x or 7.5.12.x, the
  SMF data can show a low number of TEMPSPACE1 "in use" rather
  than the value specified in the json configuration file or
  calculated due to the setting "automatic". This low value is an
  indication that the Accelerator is not working with a tempspace
  as defined in the json configuration file but operates as if the
  
  tempspace were unlimited.
  In case of query outliers (run away-queries), this processing
  may lead to a complete use of the data pool which then would
  result in a system outage.
  
  A fix for this issue will be made available with Accelerator
  maintenance level 7.5.13.
  
  Additional keywords:
  TS016108578 json configuration temp_working_space SMS DMS
  unlimited automatic DT390979
  

Local fix

Problem summary

• Problem Sumamry:
  See APAR Error description.
  
  Users Affected:
  Users having set `"temp_working_space"` to either a fixed value
  or to `"automatic"` in the json configuration file. Note that
  `"automatic"` is the default setting.
  
  
  Problem Scenario:
  See APAR Error description.
  
  Problem Symptoms:
  The disk space intended for the temporary data is not taken
  because the "temp_working_space" setting has been reverted to
  "unlimited". If an `ssh` login is possible, the command
  
  `db2pd -db bludb -tablespace -allmembers | less`
  
  will show the `TEMPSPACE1` setting on all MLNs.
  
  Its type should be DMS if "automatic" or a fixed value has been
  specified for parameter "temp_working_space".
  

Problem conclusion

• This issue has been fixed with Accelerator maintenance level
  7.5.13.
  
  Upgrade your Accelerator environment(s) accordingly. No further
  measures are necessary.
  

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:

Fix information

Applicable component levels