Using the REXX trigger

This topic provides rules and guidelines for using the REXX trigger.

REXX trigger script rules

To ensure data integrity and command execution can always completed in reasonable time, the REXX scripts developed by customers should follow the following rules:
  • A customized REXX trigger must be able to return in certain time. Customized REXX codes should not contain infinite loop or deadlock that will block following TSOC command execution.
  • A customized REXX trigger must return a result in integer format. Customized REXX codes should return a value to notify TSOC of the result. If the trigger fails, TSOC will stop and show the failure to allow the user to check what happened.
  • A customized REXX trigger can not modify files which are controlled by ClearCase. Because TSOC lacks the protection mechanism on its data, REXX codes can do anything on the data (ClearCase files). This is very dangerous and can easily break data integrity and block TSOC command execution; therefore, customized REXX codes should not modify any files controlled by ClearCase. Customized REXX scripts can change files in the “(unreserved) checked out state for files not controlled by ClearCase.
  • A customized REXX trigger must not lock any resource that will be used by a possible TSOC operation.

Interface of REXX trigger

  1. TSOC passes current command, invocation point and file name, including current PDS, to REXX trigger.
  2. The REXX trigger returns results in an integer format: 0 indicates success, other values indicate failure.