[
https://issues.apache.org/jira/browse/YARN-9173?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16735506#comment-16735506
]
liyakun commented on YARN-9173:
-------------------------------
[~wilfreds] I have read your patch, it LGTM.
The goal of YARN-8833 is to fix the infinite loops problem encountered in a
production environment, and the UT is designed so. But it does not perfectly
solve the problem caused by integer overflow. As you mentioned, `totalMaxShare`
may also beyond INT.MAX_VALUE, which is not considered by YARN-8833, and this
may cause correct issue.
> FairShare calculation broken for large values after YARN-8833
> -------------------------------------------------------------
>
> Key: YARN-9173
> URL: https://issues.apache.org/jira/browse/YARN-9173
> Project: Hadoop YARN
> Issue Type: Bug
> Components: fairscheduler
> Affects Versions: 3.3.0
> Reporter: Wilfred Spiegelenburg
> Assignee: Wilfred Spiegelenburg
> Priority: Major
> Attachments: YARN-9173.001.patch, YARN-9173.002.patch
>
>
> After the fix for the infinite loop in YARN-8833 we now get the wrong values
> back for fairshare calculations under certain circumstances. The current
> implementation works when the total resource is smaller than Integer.MAXVALUE.
> When the total resource goes above that value the number of iterations is not
> enough to converge to the correct value.
> The new test {{testResourceUsedWithWeightToResourceRatio()}} only checks that
> the calculation does not hang but does not check the outcome of the
> calculation.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
