[
https://issues.apache.org/jira/browse/MAPREDUCE-3353?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13228739#comment-13228739
]
Arun C Murthy commented on MAPREDUCE-3353:
------------------------------------------
Looks good, some minor nits:
# RegisterApplicationMasterResponse.(get,set)UnusableNodes is not used right
now, so let's add it later when we need it.
# AMResponse.getUpdatedNodes could use javadocs.
# BuilderUtils.createNodeReport
# RMContextImpl.applications shud be changed to ConcurrentSkipListMap to be
safe - we should open a separate jira to fix the signature of RMContext.get*
which return ConcurrentMap
# RMAppImpl.pullRMNodeUpdates needs a writeLock since it's clearing
# RMAppImpl shud ignore NodeUpdate in COMPLETED state (thus we can remove the
'if' condition in RMAppNodeUpdateTransition).
> Need a RM->AM channel to inform AMs about faulty/unhealthy/lost nodes
> ---------------------------------------------------------------------
>
> Key: MAPREDUCE-3353
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3353
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Components: applicationmaster, mrv2, resourcemanager
> Affects Versions: 0.23.0
> Reporter: Vinod Kumar Vavilapalli
> Assignee: Bikas Saha
> Fix For: 0.23.2
>
> Attachments: MAPREDUCE-3353-branch-0.23.patch,
> MAPREDUCE-3353-branch-0.23.patch, MAPREDUCE-3353-branch-0.23.patch,
> MAPREDUCE-3353-branch-0.23.patch
>
>
> When a node gets lost or turns faulty, AM needs to know about that event so
> that it can take some action like for e.g. re-executing map tasks whose
> intermediate output live on that faulty node.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
