[
https://issues.apache.org/jira/browse/SAMZA-893?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15196550#comment-15196550
]
Jake Maes commented on SAMZA-893:
---------------------------------
[~navina] quite the opposite. The patch for SAMZA-867 very obviously broke the
expiration logic with this line:
--- if (!expired || allocatedContainers == null || allocatedContainers.size()
== 0) {
+++ if (expired || !hasAllocatedContainer(ANY_HOST)) {
The current patch definitely fixes that regression.
Because the original unit test never caught the regression, I also tried to
improve the unit test, but it appears that was unsuccessful.
However, I don't have the cycles right now to investigate the broken unit
tests, as that was a pre-existing condition and will likely require
considerable time.
> Fix a bug with host affinity request expiration introduced in SAMZA-867
> -----------------------------------------------------------------------
>
> Key: SAMZA-893
> URL: https://issues.apache.org/jira/browse/SAMZA-893
> Project: Samza
> Issue Type: Bug
> Reporter: Jake Maes
> Assignee: Jake Maes
> Attachments: SAMZA-893.patch
>
>
> The expiration logic change in the patch for SAMZA-867 was wrong. It
> simplified the conditional, but would cause expired requests to never get
> rescheduled on ANY_HOST.
> This ticket is to fix that logic and the unit tests, which didn't fail after
> the change.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
