When the system logger becomes hung up on an operation or waits for a resource, bottlenecks can occur on that system. Because the system logger manages sysplex resources, a bottleneck on a single system can have an effect on the entire sysplex. Interrupting a particular stream request for a resource might allow much other work to continue in the system and sysplex.

The system logger monitors its allocation and HSM recall service tasks for delays and provides a mechanism (through WTORs) to interrupt these delayed requests. To the system logger, the interruption is an error condition for the current request. Removal of a delayed request enables the processing of other log stream resource requests. Messages IXG271I and IXG272E are issued if the system logger detects that a delay of a service request is inhibiting other log stream resource requests.

You can reply FAIL to message IXG272E to interrupt a delayed logger service request. This reply causes the request to fail and might also allow other work that was waiting to continue.

Use the FAIL option only if you cannot determine why the request is not completing. Replying FAIL might cause undesirable results. This option is meant to keep the rest of the system logger applications running, at the expense of one hung application.

Examine any other error messages from the system logger or from any exploiter of the affected log stream. If you reply FAIL, the system logger might cause other components such as allocation to enter their recovery, create memory dumps, or issue various messages. If you reply FAIL to message IXG272E, you might see a dump in media manager, dynamic allocation, or catalog, depending on what type of I/O delay that the system logger was experiencing at the time IXG272E was issued. There is an unexpected dump that results in abend 0E0 during SVCDMP processing.