The RM's admin service exposes a method refreshQueues that is used to update the queue configuration when used with the CapacityScheduler, but if it is used with the FifoScheduler, it will set the containerTokenSecretManager/clusterTracker fields on the FifoScheduler to null, which eventually leads to NPE. Since the FifoScheduler only has one queue that cannot be refreshed, the correct behavior is for the refreshQueues call to be a no-op.
I will attach a patch that fixes this by splitting the ResourceScheduler's reinitialize method into separate initialize/updateQueues methods.
|Field||Original Value||New Value|
|Attachment||MAPREDUCE-2630.patch [ 12484600 ]|
|Assignee||Josh Wills [ jwills ]|
|Status||Open [ 1 ]||In Progress [ 3 ]|
|Status||In Progress [ 3 ]||Patch Available [ 10002 ]|
|Status||Patch Available [ 10002 ]||Resolved [ 5 ]|
|Fix Version/s||0.23.0 [ 12315570 ]|
|Resolution||Fixed [ 1 ]|
|Status||Resolved [ 5 ]||Closed [ 6 ]|
|Transition||Time In Source Status||Execution Times||Last Executer||Last Execution Date|
|13d 9h 26m||1||Josh Wills||12/Jul/11 17:05|
|46s||1||Josh Wills||12/Jul/11 17:06|
|10h 56m||1||Mahadev konar||13/Jul/11 04:02|
|124d 21h 48m||1||Arun C Murthy||15/Nov/11 00:50|