Resource type ENQUEUE

If a transient data queue has been defined as intrapartition and logically recoverable, there are further restrictions on the use of the queue by more than one task at a time (in addition to those leading to waits on resource type TDIPLOCK).

If you have a task suspended on resource type ENQUEUE, and a value of TDNQ, the task has been suspended while attempting to read, write or delete a logically recoverable queue because a required enqueue is currently held by another task.

Note: For general information about dealing with enqueue waits, see Investigating enqueue waits. Issuing a CEMT INQUIRE UOWENQ command reveals the name of the queue and whether the enqueued read or write is required by the task. If the task is enqueued against the read end of the queue, a qualifier of FROMQ is displayed on the CEMT INQUIRE UOWENQ screen. If the task is enqueued against the write end of the queue, a qualifier of TOQ is displayed on the CEMT INQUIRE UOWENQ screen.

If you want to delete a queue, both the read and the write enqueues must be obtained. No task may, therefore, read or write to a queue while a delete operation is in progress. A delete cannot proceed until any task currently reading has completed its read or any task writing has committed its changes.

In general, a wait on a resource type of ENQUEUE should not last for long unless the task owning the enqueue has been delayed. If the UOW that owns the enqueue has suffered an indoubt failure, the UOW is shunted. If the queue accessed by this UOW is defined as WAIT=YES and WAITACTION=QUEUE, the wait can last for a long period of time. To deduce if an indoubt failure has occurred:
  • Issue a CEMT INQUIRE UOWENQ command to display the name of the enqueue owner.
  • Issue a CEMT INQUIRE UOW command to see if the UOW is shunted.


dfhs1c00150.html | Timestamp icon Last updated: Monday, 5 June 2017