Details
-
Improvement
-
Status: Resolved
-
Major
-
Resolution: Won't Fix
-
None
-
None
-
None
-
None
Description
It would be good if the cluster admin could define a job tracker name that was used to form the unique part of the job ids. I think something like:
mapred.jobtracker.name = xxx
job_xxx_200808081200_00001
if it is undefined, it would default to the current behavior of just using the timestamp.