Emergency restart

With this type of start, DBCTL is restarted in the environment it was in before a DBCTL failure.

To perform an emergency restart of DBCTL, use the /ERESTART command. DL/I in-flight UOWs (that is, those that were still being processed when the failure occurred) are backed out. Committed but unwritten DEDB changes are applied to the database. Units of work that were indoubt are retained and are resolved automatically when CICS® and DBCTL are reconnected. For further guidance on how this is done, see Operations and automation in IMS product documentation. If the UOWs fail to be resolved automatically, you can use DBCTL operator commands to do so, as described in Resolving indoubt CICS DBCTL units of work manually.

If a failure in emergency restart prevents backout being completed, instead of using a COLD start, you can reattempt the emergency restart using the COLDBASE keyword on the emergency restart command. Full function DL/I databases and DEDB areas that have indoubt data or that need backout or recovery are identified and stopped. Database backout and committed DEDB updates are not done. You must then use the appropriate IMS utilities to backout or forward recover these databases. (See Database utilities in IMS product documentation for guidance on using the utilities.)

You can also specify whether the restart or write ahead data sets should be formatted as part of the restart process. Format the RDS and the WADS if there has been a data set I/O error or if you need to reallocate a data set or change its size.