CCTL performance: monitoring DRA thread TCBs
You can evaluate the status of DRA thread TCBs from the output of the /DISPLAY CCTL ALL command in most cases. If there are no thread failures, the output might show fewer thread TCBs than the MINTHRD value because of internal short lived conditions.
- An abend.
- A nonzero DRA thread request return code that causes the thread TCB to collapse.
- Termination using a IMS DB /STOP REGION command.
Failed thread TCBs are not automatically recreated. The thread TCB number increases again if a new thread is created to process a SCHED request. If the number of thread TCBs is greater than the MINTHRD value and all thread activity ceases normally, the number of thread TCBs left in the DRA will be the MINTHRDD value.
During CCTL processing, the number of active DRA threads occupying thread TCBs varies from 0 to the MAXTHRD number. Active DRA threads indicate that at least one SCHED request has been made but not any TERMTHRD requests. If the number of non-active thread TCBs becomes too large, the DRA automatically collapses some thread TCBs to release IMS™ DB resources.
The status of DRA thread TCBs can be evaluated from the output of the /DISPLAY CCTL ALL command, except for one case.
If there were no thread failures, the output might show fewer thread TCBs than the MINTHRD value because of internal short lived conditions. In fact, the actual number of thread TCBs equals the MINTHRD.
- DRA thread statistics
DRA thread statistics are returned for a SCHED request and for any DRA requests that terminate a UOR. The statistics are in a CCTL area that is pointed to by the PAPLSTAT field. - DRA statistics
DRA statistics are contained in the returned PAPL as a result of a DRA TERM request, or in the Control exit routine's PAPL when it is called for DRA termination. This routine is called when the DRA fails or when a previous Control exit routine invocation resulted in return code 4. - DRA tracing
Tracing (logging) of activity does not occur in the DRA, but there is tracing in IMS DB of DL/I and Fast Path activity. The setup and invocation of DL/I tracing for IMS DB is the same as for IMS. The output trace records for CCTL threads contain the recovery token. - Sending commands to IMS DB
In an IMS DB warm standby or IMS XRF environment, a CCTL can use a z/OS SVC 34 to broadcast an emergency restart command to an IMS DB alternate, or a SWITCH command to an IMS XRF alternate in order to have the IMS alternate system become the primary IMS system. - Problem diagnosis
Diagnostic information is provided by the DRA in the form of an SDUMP, or a SNAP data set output. For X'80', the SDUMP is attempted first. If it fails, SNAP is done.
