[
https://issues.apache.org/jira/browse/YARN-10072?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17009806#comment-17009806
]
Jim Brennan commented on YARN-10072:
------------------------------------
I resolved this internally by changing
TestCSAllocateCustomResource#testCapacitySchedulerInitWithCustomResourceType to
use MockRM like the other test in TestCSAllocateCustomResource. It was using
a lot of mocking/spying to try to isolate CapacityScheduler, but in so doing I
think it introduced some inconsistency in the initialization process - I was
getting inconsistent results depending on where I set breakpoints while
debugging. By using MockRM, the CapacityScheduler initialization should more
closely match what happens in production. And it removed the inconsistency
with breakpoints.
I will put up a patch for trunk shortly.
> TestCSAllocateCustomResource failures
> -------------------------------------
>
> Key: YARN-10072
> URL: https://issues.apache.org/jira/browse/YARN-10072
> Project: Hadoop YARN
> Issue Type: Test
> Components: yarn
> Affects Versions: 2.10.0
> Reporter: Jim Brennan
> Priority: Major
>
> This test is failing for us consistently in our internal 2.10 based branch.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
