[
https://issues.apache.org/jira/browse/MAPREDUCE-2876?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13091820#comment-13091820
]
Robert Joseph Evans commented on MAPREDUCE-2876:
------------------------------------------------
The issue is with granularity and how far off from the expire interval the
expire will actually be. In the worst case something will expire just under 1
monitor interval after it should have. The only issue would be if someone
really needed that expiry to be much closer to the actual expire interval then
1/3 of it. I don't really see that as much of an issue for anything we have
right now.
I would propose that we modify LivnessMonitor to store expiry in a priority
queue instead of a map. Then it can sleep just the amount of time needed
before the next one expires. It is fairly simple and I have done this type of
thing in the past. Then there is very little error from the expire interval,
and we don't have to worry about a monitor interval at all.
> ContainerAllocationExpirer appears to use the incorrect configs
> ---------------------------------------------------------------
>
> Key: MAPREDUCE-2876
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-2876
> Project: Hadoop Map/Reduce
> Issue Type: Bug
> Components: mrv2
> Affects Versions: 0.23.0
> Reporter: Robert Joseph Evans
> Assignee: Robert Joseph Evans
> Fix For: 0.23.0
>
>
> ContainerAllocationExpirer sets the expiration interval to be
> RMConfig.CONTAINER_LIVELINESS_MONITORING_INTERVAL but uses
> AMLIVELINESS_MONITORING_INTERVAL as the interval. This is very different
> from what AMLivelinessMonitor does.
> There should be two configs RMConfig.CONTAINER_LIVELINESS_MONITORING_INTERVAL
> for the monitoring interval and RMConfig.CONTAINER_EXPIRY_INTERVAL for the
> expiry.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
