CICS keypoints and IMS checkpoints

This section discusses system-level keypoint and checkpoint information. Both CICS® and IMS also have task or program (thread) level synchronization information.

CICS keypoints and IMS checkpoints both contain system status information that is modified during online operation. The concepts are basically the same, but they are implemented differently.

A CICS warm start uses a warm keypoint that was written to the CICS catalog by the previous normal CICS shutdown.

A CICS emergency restart reads the CICS system log backwards until it has located an activity keypoint. The keypoint contains a record of incomplete UOW chains which CICS reads directly. These chains can reside on the primary and secondary system logs.

An IMS warm start reads the checkpoint ID table on the RDS to find the shutdown checkpoint on the log. The RDS is a data set that IMS uses to record system checkpoint ID information during the logging process. IMS finds the information it needs and uses it automatically. If the RDS is not available at restart, you can obtain the checkpoint information needed from the log, but this may lengthen the restart process. Generally, you do not need to know the content of the RDS. However, if you are faced with a particularly complex recovery problem, you may need to examine the RDS. You can find guidance on its contents in Operations and automation in IMS product documentation.

An IMS emergency restart reads the checkpoint ID table on the RDS and selects the checkpoint that precedes the last synchronization point of each program that was active at the time of the failure. It then reads the IMS log forward from the selected checkpoint.

To take a simple checkpoint of DBCTL, use the /CHECKPOINT command.