CICS has stalled during initialization

If CICS stalls during initialization, on an initial, cold, warm, or emergency start, the first place to look is the MVS™ console log. This tells you how far initialization has progressed.

Note that there might be significant delays at specific stages of initialization, depending on how CICS last terminated.

On a cold start, loading the GRPLIST definitions from the CSD data set can take several minutes. For large systems, the delay could be 20 minutes or more while this takes place. You can tell if this stage of initialization has been reached because you get this console message:
DFHSI1511 INSTALLING GROUP LIST xxxxxxxx

On a warm start, there may be a considerable delay while resource definitions are being created from the global catalog.

If you find that unexpected delays occur at other times during CICS initialization, consider the messages that have already been sent to the console and see if they suggest the reason for the wait. For example, a shortage of storage is one of the most common causes of stalling, and is always accompanied by a message. The JCL job log is another useful source of information.

You can find out if this has happened by taking an SDUMP of the CICS region. Format the dump using the keywords KE and DS, to get the kernel and dispatcher task summaries.

Consider, too, whether any first-or second-stage program list table (PLT) program that you have written could be in error. If such a program does not follow the strict protocols that are required, it can cause CICS to stall. For programming information about PLT programs, see Writing initialization and shutdown programs.



dfhs135.html | Timestamp icon Last updated: Monday, 5 June 2017