Sequencer restart/resume job: Handle uncommited records

[AKUMAR21]

I have a sequence job that runs 25 jobs in parallel. Each of the 25 jobs uses separate MQ stage that destructivley reads records from separate MQ and loads into Teradata through TPUMP stage. In case a job aborts, how to handle those records that were destructively read from MQ and that failed to commit to Teadata? Can we restart/resume the job from sequencer without losing any data?

[ArndW]

I would add writing the MQ-Series record to a sequential file so that it isn't lost should the job abort. You cannot restart the sequence without losing data unless you explicitly program storing the interim data.