Recovery actions

If the amount of storage that is being monitored continues to grow, determine whether it is expected behavior. Use the available RRS data collection techniques such as panels, console display command, or batch program. These techniques can help you assess the level of transaction activity in RRS and determine if it is unusual or unexpected. The threshold values can be updated for each IBM® Health Checker for z/OS® check to values more appropriate to the environment. Follow the action based on:

• RRS_Storage_NumLargeMSGBlks exception messages:
  • Look for hangs with sysplex cascaded transactions and base actions on the hang reported
  • Use the DISPLAY RRS,UREX command to list all active units of recovery (UR) waiting for an event.
  • If possible, resolve the hang. Otherwise, capture a dump of RRS and the associated resource manager to provide to the IBM Support Center for further assistance.
• RRS_Storage_NumTransBlks exception messages:
  • Enter the DISPLAY RRS,UR command to list all active UR.
  • Investigate which application is creating these URs and check for a problem with that application, such as looping, not cleaning up transactions, or recursive errors.
  • If you do not understand the increase in transaction activity:
    • Collect a dump of RRS and any resource manager or work manager identified as the source of the increased transaction activity.
    • Provide the dump to the IBM Support Center for further assistance.
• For any high storage count:
  • Look for the RRS_Storage_NumServerReqs check exception message, ATRH016E. The message suggests a possible hang in RRS processing, which can cause storage usage to grow.
  • Capture a dump of RRS and provide it to the IBM Support Center for further assistance.