Subsystem inactivity management

If activity slows down in participating IMS subsystems, ensure that ordering operations can continue in your replication environment.

In an IMS data sharing environment where you are capturing changes from multiple DB/DC or DBCTL subsystems, one or more subsystems can stop producing new log records for an extended period of time. Data Replication for IMS cannot remove inactive subsystems from ordering decisions, because the source server must receive activity from all of them to avoid replicating changes out of order.

In some cases, an IMS subsystem does not produce enough records to fill a buffer, and does not physically write the data to the online log data sets (OLDS) until the buffer is full. However, you can use automation or manual intervention to flush the buffers and ensure that the source server receives a steady feed of changes from each participating subsystem.

To help you address this requirement, the source server issues the CECZ0400W message to identify the subsystem that became inactive. The user-configurable parameter INACTTHRESHOLD determines how frequently this message is displayed.