Restart with a single failing IRLM

If the only IRLM in a data-sharing configuration fails: the associated IMS online systems suspend all transaction scheduling, IMS performs backout for affected transactions, IMS places the transactions on a suspend queue or transaction queue, depending on whether you specified SERIAL=NO or SERIAL=YES in the TRANSACT macro and IMS stops shared databases.

IMS issues DFS2012I messages for each database that has been stopped automatically because of the failure. IMS abnormally terminates batch IMS systems associated with the IRLM. Batch database backout is required for only those batch executions that perform database updates and do not use dynamic backout.

After you restart the failed IRLM, enter a z/OS® MODIFY command, specifying the RECONNECT keyword, for each IMS online system. This command reestablishes the data-sharing environment, and initiates reprocessing for the suspend queue.

Related reading: For more information about transaction recovery, see Transaction recovery in online IMS systems)