Analyzing DRA problems

To analyze DRA problems, first investigate any external conditions that might have caused the problem. If you can eliminate external causes, an unexpected DBCTL return code or another IMS function might have caused the problem.

1. Did external conditions cause the problem?
   • For CCTL external problems, check the status of applications or transactions. DBCTL and the DRA do not control these resources.
   • For DBCTL external problems, check the status of databases, PSBs, and dependent regions (BMPs and CCTLs) by using the /DISPLAY commands.
   • For DRA external problems:
     • Make sure you are using the correct DRA startup table for this DBCTL/CCTL session. Values such as Fast Path buffer allocations and minimum/maximum thread specifications can cause scheduling and resource problems.
     • Become familiar with the CCTL control exit.

       The DRA calls the control exit to notify the CCTL of certain events, such as a DRA failure, an identify failure, a DBCTL failure, and so on. The DRA passes this information in a parameter list (DFSPAPL). The CCTL responds by passing back a return code in field PAPLRETC to tell the DRA what action to perform. Understanding which actions the CCTL is allowed to request can help you distinguish between valid actions and failures.

     • The DRA does not issue any messages that report the actions it performed.
   • If an external condition caused the problem, stop here and fix the problem. Otherwise, continue with the next step.
2. You reach this point by eliminating external reasons as the cause of the problem.
   Determine if DBCTL returned a nonzero return code, indicating that the request from the CCTL was not successfully completed.

   • If yes, take a z/OS® online dump of the CCTL and contact IBM® Software Support.
   • If no, then other functions might be involved in the problem. Use the appropriate section in this information to analyze the problem. Keyword procedures are useful in narrowing the problem to a specific cause.