A customer stores components in datasets with a logical record length of 230 bytes (or greater) and uses Stacked Reverse Deltas ( SRDs ) to house this data in their baseline libraries. In ZMF 5.6.0, two distinct problems have been reported so far.
ZMF 7.1.2, when ERO audit performed version regression checks it compared the statistics for the component in the area being audited against statistics for the same component which had been “first found” in the consolidated baseline search concatenation. This search concatenation consisted of higher areas in the current release, then areas in the prior releases and finally in the baseline libraries.
This KB article contains a fix for the problem that should be applied to all ChangeMan ZMF 8.1.1 systems in a timely manner and before implementation of 8.1.1 into a live environment wherever possible.
This problem may occur when the same baseline dataset is shared either between multiple applications or between multiple library types within a single application.
Alternatively customers can run the IMPACT & LDSLOAD jobs on the same schedule. This has the added advantage that it will both re-organize the IA dataspace and also ensure that any small changes to baseline library definitions that have been made but forgotten about will also be picked up and reflected in the IA table contents. This is something that we often see here and usually presents itself in the form of unexpected or missing SYNCH errors in audit jobs.
