Recovery

Any software program runs the risk of interruption from a hardware or software event that unexpectedly stops execution. The InterChange Server has a robust mechanism for recovering transactions that are in progress when an unexpected exit occurs.

When the InterChange Server comes back up after an unexpected exit, it checks for collaborations that were in an active transaction state at the time of exit. A two-phased recovery then begins:

  1. The InterChange Server reactivates each interrupted transactional collaboration and rolls it back (unless the collaboration was using the long-lived business process feature). During this period, the server does not deliver new events to the collaboration.
  2. The InterChange Server retrieves the original triggering event for each interrupted collaboration from the event management service and redelivers it. The collaboration runs, reprocessing the event.

When recovery is complete, the InterChange Server allows the collaboration to process new events.

Copyright IBM Corp. 1997, 2004