[
https://issues.apache.org/jira/browse/MAPREDUCE-3186?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13136648#comment-13136648
]
Eric Payne commented on MAPREDUCE-3186:
---------------------------------------
@Mahadev
Would you be able to suggest the best way to unit test this?
Here are the manual tests I performed:
# I ran the following in a single node cluster:
## Start the {{wordcount}} test with a 1.5G file
## Make sure the {{wordcount}} application is in the RUNNING state.
## Stop the RM, HS, and NM daemons
## Restart the RM, HS, and NM daemons
## Check that MRAppMaster java task has exited and log contains the following
string: "ERROR [RMCommunicator Allocator]
org.apache.hadoop.mapreduce.v2.app.rm.RMContainerAllocator: Error communicating
with RM: Resource Manager doesn't recognize AttemptId:
application_XXXXXXXXXXXXX_XXXX"
# I ran the following in a single node cluster:
## Start the {{wordcount}} test with a 1.5G file
## Make sure the {{wordcount}} application is in the RUNNING state.
## Stop the RM, HS, and NM daemons
## Wait for the MRAppMaster java task to exit. This may take a few minutes.
## Check that MRAppMaster java task has exited and log contains the following
string: "ERROR [RMCommunicator Allocator]
org.apache.hadoop.mapreduce.v2.app.rm.RMContainerAllocator: Error communicating
with RM: Could not contact RM after 10 tries."
## 10 is the default number of retries.
# I ran the following in a single node cluster:
## Edit the yarn-site.xml config file.
## Add the following property: <br> name =
yarn.app.mapreduce.am.scheduler.connection.retries <br> value = 7
## Start the {{wordcount}} test with a 1.5G file
## Make sure the {{wordcount}} application is in the RUNNING state.
## Stop the RM, HS, and NM daemons
## Wait for the MRAppMaster java task to exit. This may take a few minutes.
## Check that MRAppMaster java task has exited and log contains the following
string: "ERROR [RMCommunicator Allocator]
org.apache.hadoop.mapreduce.v2.app.rm.RMContainerAllocator: Error communicating
with RM: Could not contact RM after 7 tries."
> User jobs are getting hanged if the Resource manager process goes down and
> comes up while job is getting executed.
> ------------------------------------------------------------------------------------------------------------------
>
> Key: MAPREDUCE-3186
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-3186
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Components: mrv2
> Affects Versions: 0.23.0
> Environment: linux
> Reporter: Ramgopal N
> Assignee: Eric Payne
> Priority: Blocker
> Labels: test
> Attachments: MAPREDUCE-3186.v1.txt
>
>
> If the resource manager is restarted while the job execution is in progress,
> the job is getting hanged.
> UI shows the job as running.
> In the RM log, it is throwing an error "ERROR
> org.apache.hadoop.yarn.server.resourcemanager.ApplicationMasterService:
> AppAttemptId doesnt exist in cache appattempt_1318579738195_0004_000001"
> In the console MRAppMaster and Runjar processes are not getting killed
--
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
