drill-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Deneche A. Hakim (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (DRILL-3147) tpcds-sf1-parquet query 73 causes memory leak
Date Wed, 20 May 2015 23:28:59 GMT

    [ https://issues.apache.org/jira/browse/DRILL-3147?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14553338#comment-14553338
] 

Deneche A. Hakim commented on DRILL-3147:
-----------------------------------------

For this query, a HashJoin builds an empty hash table and thus it can finish the query right
away. It will kill it left stream which will send a "receiver finished" message to all it's
senders, and then close it's resources (including it's allocator).
If at the same time, a batch has just been received from one of the senders, sometimes the
fragment's allocator will close itself before the rpc layer (DataServer) had time to release
all references to the batch's buffer

> tpcds-sf1-parquet query 73 causes memory leak
> ---------------------------------------------
>
>                 Key: DRILL-3147
>                 URL: https://issues.apache.org/jira/browse/DRILL-3147
>             Project: Apache Drill
>          Issue Type: Bug
>            Reporter: Deneche A. Hakim
>            Assignee: Deneche A. Hakim
>             Fix For: 1.1.0
>
>
> the leak seems to appear when BaseRawBatchBuffer.enqueue() tries to release a batch but
the allocator has already been closed 



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Mime
View raw message