Why is there a discrepancy between the date time stamp on messages DC001003, DC201006 and DC027007 written to the IDMS CV log and the corresponding DC410003 and DC410006 written to the batch log when an external run unit aborts?
Whenever the external wait time is exceeded for an external run unit, messages DC001003 STALLED WAITING FOR EREECB, DC201006 CV-Status BE-TaskID Pri FE - ID1 FE and DC027007 PROGRAM ABENDED WITH CODE D002 are written to the IDMS CV LOG (JES2.JESMSGLG).
Corresponding to these messages, a DC410003 DNS SEND/RECEIVE Failure Status is: 5839 followed by a DC410006 DNS Processing error, Function is RECEIVE_AND_WAIT are written to the Batch log (JES2.JESMSGLG). There is usually a difference in time these messages are written to the IDMS CV log and the Batch log suggesting that they may not be
The batch region starts an external run unit with IDMS and accesses a database. Leaving the run unit open it continues processing without returning to the run unit with another DML request until after the run unit is terminated by IDMS because it exceeded the external wait time.
After the run unit has been opened for longer than the external wait time without receiving another DML request from the batch job, IDMS terminates the run unit and writes DC001003, DC201006 and DC027007 to the IDMS CV log. It puts a date time stamp for the moment in which this abort of the run unit occurs on these messages.
The batch continues processing not knowing that the run unit was terminated.
When it attempts the next DML request after whatever processing it had to complete it will fail with the DC410003/5839 and DC410006 messages. These are written to the batch log with the date time stamp for the moment this abort occurs.
The difference in time is therefore how long the batch job takes to attempt another DML request after the external run unit is aborted by IDMS because the external wait was exceeded.
System Generation Guide