Jason Lowe commented on YARN-3965:

This still has utility outside of a massive polling script.  The NM provides a 
web UI just like the other daemons, reporting what hadoop version it's running, 
when it last health reported, etc.  The startup time is conspicuously missing 
compared to similar About pages on other Hadoop daemons.

Also this is not quite the same concept as when an NM joined the RM.  An NM can 
join the RM multiple times since it started, and sometimes one would like to 
know when the NM process itself launched.  Sure, if one has access they can 
login to the node and figure it out from the OS, but it can be convenient to 
report it on the nodemanager UI as well.

> Add starup timestamp for nodemanager
> ------------------------------------
>                 Key: YARN-3965
>                 URL: https://issues.apache.org/jira/browse/YARN-3965
>             Project: Hadoop YARN
>          Issue Type: Improvement
>          Components: nodemanager
>            Reporter: Hong Zhiguo
>            Assignee: Hong Zhiguo
>            Priority: Minor
> We have startup timestamp for RM already, but don't for NM.
> Sometimes cluster operator modified configuration of all nodes and kicked off 
> command to restart all NMs.  He found out it's hard for him to check whether 
> all NMs are restarted.  Actually there's always some NMs didn't restart as he 
> expected, which leads to some error later due to inconsistent configuration.
> If we have startup timestamp for NM,  the operator could easily fetch it via 
> NM webservice and find out which NM didn't restart, and take mannaul action 
> for it.

This message was sent by Atlassian JIRA

Reply via email to