[ 
https://issues.apache.org/jira/browse/YARN-7150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16151883#comment-16151883
 ] 

anikad ayman commented on YARN-7150:
------------------------------------

Thank you [~templedf] for your answer :) 

In fact, the problem is more complicated just after this message in the log, 
all the processing craches => That means the two jobs were running very slowly 
for long time ( more than 7 hours)  without incresming in the state, and 
several jobs was in waiting state => Just after killing the 2 running jobs, all 
the jobs started to run and ended very fastly. 

We had this issue several times before ( at least 1 time in every mounth )  
thing that makes a lot of problems in our production envirenment !

Do you think that increasing maxCompletedAppsInStateStore will solve the 
problem ? 



> Yarn crash [max number of completed apps kept in memory met]
> ------------------------------------------------------------
>
>                 Key: YARN-7150
>                 URL: https://issues.apache.org/jira/browse/YARN-7150
>             Project: Hadoop YARN
>          Issue Type: Bug
>         Environment: Production
>            Reporter: anikad ayman
>
> During MapReduce processing of several jobs, Yarn did crash and the 
> processing of jobs had stopped.
> I successed to back the processing after killing jobs which were running 
> (2jobs).
> In the logs I find from the beginning of the crash :
> {code:java}
> *Max number of completed apps kept in state store met*: 
> maxCompletedAppsInStateStore = 10000, removing app 
> application_1500982512144_26754 from state store.
>  2017-08-25 03:50:05,799 INFO 
> org.apache.hadoop.yarn.server.resourcemanager.RMAppManager: *Application 
> should be expired, max number of completed apps kept in memory met*: 
> maxCompletedAppsInMemory = 10000, removing app 
> application_1500982512144_26754 from memory
> {code}
> After that , this message shows up several times in the log :
> {code:java}
> Large response size 4742320 for call 
> org.apache.hadoop.yarn.api.ApplicationClientProtocolPB.getApplications
> {code}
> Have you any explication and solution of this issue ?



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to