[MAPREDUCE-7355] Fix MRAppMaster to getStagingAreaDir from Job Configuration

• Type: Improvement
• Status: Open
• Priority: Major
• Resolution: Unresolved
• Affects Version/s: 3.3.0
• Component/s: job submission
• Assignee: Prabhu Joseph
• Reporter: Prabhu Joseph
• Created: 21/Jun/21 19:58
• Updated: 23/Jun/21 15:37

Description

When JobClient (runs as yarn user) uses RM_DELEGATION_TOKEN (owner:oozie) to submit the job, client uses /mapreducestaging/yarn as staging directory whereas MRAppMaster uses /mapreducestaging/oozie. This leads to below failure

Job init failed : org.apache.hadoop.yarn.exceptions.YarnRuntimeException: java.io.FileNotFoundException: hdfs://yarncluster/mapreducestaging/oozie/.staging/job_1624284676187_0003/job.splitmetainfo: No such file or directory.
at org.apache.hadoop.mapreduce.v2.app.job.impl.JobImpl$InitTransition.createSplits(JobImpl.java:1611)
at org.apache.hadoop.mapreduce.v2.app.job.impl.JobImpl$InitTransition.transition(JobImpl.java:1473)
at org.apache.hadoop.mapreduce.v2.app.job.impl.JobImpl$InitTransition.transition(JobImpl.java:1431)
at org.apache.hadoop.yarn.state.StateMachineFactory$MultipleInternalArc.doTransition(StateMachineFactory.java:385)
at org.apache.hadoop.yarn.state.StateMachineFactory.doTransition(StateMachineFactory.java:302)
at org.apache.hadoop.yarn.state.StateMachineFactory.access$500(StateMachineFactory.java:46)
at org.apache.hadoop.yarn.state.StateMachineFactory$InternalStateMachine.doTransition(StateMachineFactory.java:487)
at org.apache.hadoop.mapreduce.v2.app.job.impl.JobImpl.handle(JobImpl.java:1010)
at org.apache.hadoop.mapreduce.v2.app.job.impl.JobImpl.handle(JobImpl.java:141)
at org.apache.hadoop.mapreduce.v2.app.MRAppMaster$JobEventDispatcher.handle(MRAppMaster.java:1544)
at org.apache.hadoop.mapreduce.v2.app.MRAppMaster.serviceStart(MRAppMaster.java:1263)

MRAppMaster can rely on Job Configuration mapreduce.job.dir to avoid this issue.