[
https://issues.apache.org/jira/browse/SLIDER-1233?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16099160#comment-16099160
]
Gour Saha commented on SLIDER-1233:
-----------------------------------
+1. The patch looks good. I successfully ran the new tests and also verified
that they fail with the absence of the src code patch.
Only one minor comment on the below src code comment -
bq. // don't increment failure counts if container was aborted or killed
I think we should not stop at _killed_, since there are other killed scenarios
like pmem and vmem which we are treating differently. Should we explicitly
specify killed for which scenarios, or maybe just shorten the comment to "//
don't increment failure counts" and folks can refer to ContainerOutcome.class
to see which scenarios count as Completed? What do you think?
I am running all the UTs now and will report back if I see any failures.
> Lost nodes should not contribute to container failures
> ------------------------------------------------------
>
> Key: SLIDER-1233
> URL: https://issues.apache.org/jira/browse/SLIDER-1233
> Project: Slider
> Issue Type: Bug
> Components: core
> Reporter: Billie Rinaldi
> Assignee: Billie Rinaldi
> Fix For: Slider 1.0.0
>
> Attachments: SLIDER-1233.001.patch
>
>
> If a container completes due to an NM being lost, we should not count this
> towards container failures that may eventually cause the AM to fail the
> application. We are already using a ContainerOutcome of Completed (rather
> than Failed) for this type of container exit, so we just need to change the
> failure counting in that case. Other failure types associated with Completed
> are killed by the AM, killed by the RM, and killed after app completion, none
> of which need to contribute to container failures.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
