[
https://issues.apache.org/jira/browse/TEZ-1269?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14059615#comment-14059615
]
Gopal V commented on TEZ-1269:
------------------------------
[~bikassaha]: this works as documented, but the containers kept are mostly the
newest containers allocated.
On one hand, it does retain containers that was node-local allocation for the
last query, but this results in somewhat sub-optimal JIT performance for short
& fast BI queries.
I'm a bit conflicted on that, but +1 for the patch, we can revisit the
container "eviction" algorithm later.
> TaskScheduler prematurely releases containers
> ---------------------------------------------
>
> Key: TEZ-1269
> URL: https://issues.apache.org/jira/browse/TEZ-1269
> Project: Apache Tez
> Issue Type: Bug
> Reporter: Bikas Saha
> Assignee: Bikas Saha
> Attachments: TEZ-1269.1.patch, TEZ-1269.2.patch
>
>
> It checks for session mode and if not true, and if there are no outstanding
> requests, then it releases the containers before the container timeout has
> expired. If the state machine is on its way to scheduling new tasks during
> this time then they will not be able to reuse these containers.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
