Database apply user exit programs cannot be used in data groups that are configured for multithreaded database apply processing. The following considerations concerning performing user journal replication of IFS objects, data areas, and data queues in a data group that has user exit programs is relevant only when the data group is configured to use single-threaded database apply processing.
When new or different journaled object types are added to an existing data group, user exit programs may be affected. Be aware of the following exit program considerations when changing an existing configuration to include IFS objects, data areas, or data queues configured for replication processing from a user journal.
-
When IFS objects, data areas, or data queues are journaled to a user journal, new journal entry codes are provided to the user exit program. If the user exit program interprets the journal code, changes may be required.
-
The path name for IFS objects cannot be interpreted in the same way as it can for database files. MIMIX uses the file ID (FID) to identify the IFS object being replicated. User exit programs that rely on the library and file names in the journal entry may need to be changed to either ignore IFS journal entries or process them by resolving the FID to a path name using the IBM-supplied APIs.
-
Journaled IFS objects and data queues can have incomplete journal entries. For incomplete journal entries, MIMIX provides two or more journal entries with duplicate journal entry sequence numbers and journal codes and types to the user exit program when the data for the incomplete entry is retrieved. Programs need to correctly handle these duplicate entries representing the single, original journal entry.
-
Journal entries for journaled IFS objects, data areas, and data queues will be routed to the user exit program. This may be a performance consideration relative to user exit program design.
Contact your Certified MIMIX Consultant for assistance with user exit programs.