hive-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Szehon Ho (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HIVE-8865) Needs to set hashTableMemoryUsage for MapJoinDesc
Date Fri, 14 Nov 2014 05:54:34 GMT

    [ https://issues.apache.org/jira/browse/HIVE-8865?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14211887#comment-14211887
] 

Szehon Ho commented on HIVE-8865:
---------------------------------

Mostly looks good.  Few nits:

 if (nd.getName().equals("GBY")) {,   We should not use the GroupByOperator.getOperatorName()?
 Also do we need this additional check given we already pass in rules?  

private void checkGroupBy(...    can have this method return boolean, and put all the context.followedByGroupBy
set/access in there?  It might be cleaner this way.

Thanks.

> Needs to set hashTableMemoryUsage for MapJoinDesc
> -------------------------------------------------
>
>                 Key: HIVE-8865
>                 URL: https://issues.apache.org/jira/browse/HIVE-8865
>             Project: Hive
>          Issue Type: Sub-task
>          Components: Spark
>    Affects Versions: spark-branch
>            Reporter: Chao
>            Assignee: Chao
>         Attachments: HIVE-8865.1-spark.patch
>
>
> If this part is not done, hashTableMemoryUsage is always 0.0, which will cause MapJoinMemoryExhaustionException.



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

Mime
View raw message