[
https://issues.apache.org/jira/browse/DRILL-1866?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14390816#comment-14390816
]
Chris Westin commented on DRILL-1866:
-------------------------------------
We've seen these caused by the client disconnecting before the query has
wrapped up its processing, and the test tries to shutdown the Drillbit
immediately, and that goes through before the query processing has been cleaned
up -- the allocators complain because the query cleanup hasn't had a chance to
close them yet. There's code in Drillbit now that won't allow Drillbit.close()
to proceed until all currently executing fragments complete; that stopped this
from happening. So this probably just needs to be verified.
> Tests that include limit sporadically fail when run as part of entire test
> suite on Linux
> -----------------------------------------------------------------------------------------
>
> Key: DRILL-1866
> URL: https://issues.apache.org/jira/browse/DRILL-1866
> Project: Apache Drill
> Issue Type: Bug
> Components: Tools, Build & Test
> Reporter: Jacques Nadeau
> Assignee: Steven Phillips
> Priority: Critical
> Fix For: 0.9.0
>
>
> Seems to be a timing issue where memory is not being released as part of
> limit cancellation of a query.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
