CQS recovery functions

CQS provides functions for recovering work-in-progress, queues, and resources in case of system shutdown or failure. Some of these recovery functions are built-in and do not require intervention.

Other functions are optional and can be set up or initiated as you need them.

System checkpoint

To enable CQS restart in the event of failure, CQS periodically takes a snapshot of all control blocks and tables, and writes that information to the z/OS® log. That process is called system checkpoint. System checkpoint can be initiated by CQS, the client, or manually with an IMS command.

CQS logging and the z/OS system logger

CQS uses the z/OS system logger to record information necessary for CQS to recover queue structures and restart. CQS writes log records for each recoverable coupling facility list structure pair that it uses to a separate log stream. The log stream is shared among all CQS address spaces that share the structure. The system logger provides a merged log for all CQS address spaces that are sharing queues on a coupling facility list structure.

Important: Changes to resource structures and nonrecoverable (RECOVERABLE=NO) queue structures are not logged.

Structure checkpoint

To enable queue structure recovery in case of failure, CQS periodically takes a ”snapshot” of the queues on all queue structures. That process is called structure checkpoint. Structure checkpoint can be initiated by CQS, the client, or manually with an IMS command.

Important: Structure checkpoint is not supported for resource structures.

CQS client requests

CQS client systems communicate with CQS using a general use interface consisting of CQS requests.