Rohith Sharma K S commented on YARN-3999:

thanks [~jianhe] for the patch, I have some doubts/comments on the patch..
bq. If this time becomes larger than 10 mins, all applications will expire
For expire , the new events would trigger which is blocked since dispatcher is 
draining. Am I missing any specific scenario which can occure?
# Why the end time considering half of the rm-expire interval? just to know any 
feasibility reason. 
# Proceding to NM-RM expire, now gracefull shutdown of NM is also supported. So 
IIUC NM expire can also trigger if draining is taking more than 10 minutes. 
Thinking any impact on unregister  nodemanager? Now timeout is based on the 
AM_RM expire config, IAC, should NM-RM expire also considered?any thoughts?
#  Ther is issue exist for test failures in TestRMAdminService, so I think 
changes for JvmPauseMonitor.java  can be tracked in separate jira i.e. 
YARN-4035. What do you think?

> Add a timeout when drain the dispatcher
> ---------------------------------------
>                 Key: YARN-3999
>                 URL: https://issues.apache.org/jira/browse/YARN-3999
>             Project: Hadoop YARN
>          Issue Type: Bug
>            Reporter: Jian He
>            Assignee: Jian He
>         Attachments: YARN-3999.1.patch, YARN-3999.2.patch, YARN-3999.2.patch, 
> YARN-3999.patch, YARN-3999.patch
> If external systems like ATS, or ZK becomes very slow, draining all the 
> events take a lot of time. If this time becomes larger than 10 mins, all 
> applications will expire. We can add a timeout and stop the dispatcher even 
> if not all events are drained.

This message was sent by Atlassian JIRA

Reply via email to