The various tasks of creating the Q replication control tables, Replication Queue Maps, and Q subscriptions will populate a number of the control tables. The following figure shows which tables are populated for bidirectional replication:
In addition to the control tables being populated at the Q replication definition stage, some are also populated during the normal operation of Q Capture and Q Apply.
Many of the control tables, for example IBMQREP_CAPMON
and IBMQREP_APPLYMON
are updated on a regular basis by Q Capture and Q Apply. For bidirectional replication, the IBMQREP_SIGNAL
table can grow very large. There has to be some mechanism by which the amount of data in these tables is controlled. This mechanism is called pruning and is covered in the next section.