The IBM BAW application runs on top of IBM WebSphere Application Server (WAS). In a typical BAW installation, there are three JVMs-AppClusterMember, MEClusterMember, and SupClusterMember-deployed on a single node, with the Messaging Engine (ME) being one of them. When the database connection is lost - due to a shutdown, restart, or network issues - the Messaging Engine typically stops automatically after a short period. You can verify this in WAS Administrative Console at Service integration -> Buses -> bus_name -> [Topology] Messaging engines -> engine_name (will be shown as stopped).
After the database becomes available again, the Messaging Engine remains stopped, requiring manual intervention to restart it. Even after restarting the Messaging Engine, the IBM BAW portals (such as Process Portal and Process Admin) remain unresponsive until all JVMs are stopped and restarted, which disrupts operations.
However, the following configurations can help maintain the Messaging Engine in a running state or automatically restart it if the database connection is lost. With these settings in place, once the database is available again (within the configured retry interval, approximately 1 hour and 6 minutes in example below), the Messaging Engine will automatically recover-allowing IBM BAW applications and portals to resume normal functionality without requiring manual action from an administrator (to start the Messaging Engine manually).
In Service integration -> Buses -> bus_name -> [Topology] Messaging engines -> engine_name -> Custom Properties (under Additional Properties section), add below custom propoerties:
sib.meAutoReenablePeriod = 40000
sib.meReenableCount = 100
After applying the above configurations, restart all the JVMs to ensure the changes take effect. Once completed, simulate a database outage and observe the behavior after the database is back online. The IBM BAW portals should resume functioning normally without the need to restart the JVMs.
These configurations allow the system to continuously retry database connections for up to approximately 1 hour and 6 minutes. During this period, the Messaging Engine will attempt to reconnect automatically, ensuring that IBM BAW portals can recover and operate normally once the database becomes available - eliminating the need for manual JVMs restarts.
Above configurations are tried and tested on an IBM BAW version 21.0.2
If you have any suggestions or better idea on this topic, you are highly encouraged to share.
------------------------------
Muhammad Haris Khan
------------------------------