System failures

These errors may occur due to a hardware or software failure that causes the integration bus to terminate.

Recovery from such an occurrence is managed by the XA coordinated transaction support of the underlying middleware and database.

In the case of a system failure, any database updates and IBM® MQ queue reads or writes that have been performed by any incomplete message flows are automatically rolled back. When the integration node is restarted, all rolled back activities are restarted.

This recovery support does not require any special logic to be built for FTM. However, it requires that all message flows be transactional and all IBM MQ queues be persistent.

If XA coordination cannot be used and the CORE value table entry 'RENDER_EVENTS_TO_DB' is set to 'Y', any events associated with an object that has been recovered will be retrieved from the database and re-issued. It should be noted that there is a performance hit with this setting and using XA is the recommended approach.