Some objects, such as data areas and data queues can have significant activity against them and can cause MIMIX to use significant CPU resource.
One or several programs can use the QSNDDTAQ and QRCVDTAQ APIs to generate thousands of journal entries for a single *DTAQ. For each journal entry, system journal replication processes package all of the entries of the *DTAQ and sends it to the apply system. MIMIX then individually applies each *DTAQ entry using the QSNDDTAQ API.
If the data group is configured for multiple Object retrieve processing (OBJRTVPRC) jobs, then several object retrieve jobs could be started (up to the maximum configured) to handle the activity against the *DTAQ.
MIMIX contains redundancy logic that eliminates multiple journal entries for the same object when the entire object is replicated. When you configure a data group for system journal replication, you should:
-
Place all *DTAQs in the same object-only data group
-
Limit the maximum number of object retrieve jobs for the data group to one. Defaults can be used for the other object data group jobs.