hadoop-hive-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Paul Yang (JIRA)" <j...@apache.org>
Subject [jira] Commented: (HIVE-1115) optimize combinehiveinputformat in presence of many partitions
Date Tue, 27 Apr 2010 19:56:32 GMT

    [ https://issues.apache.org/jira/browse/HIVE-1115?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12861514#action_12861514
] 

Paul Yang commented on HIVE-1115:
---------------------------------

Not quite sure if this will be completed this quarter - I believe MAPREDUCE-1501 was helpful
in handling a similar use case?

> optimize combinehiveinputformat in presence of many partitions
> --------------------------------------------------------------
>
>                 Key: HIVE-1115
>                 URL: https://issues.apache.org/jira/browse/HIVE-1115
>             Project: Hadoop Hive
>          Issue Type: Improvement
>          Components: Query Processor
>            Reporter: Namit Jain
>            Assignee: Paul Yang
>
> A query like :
> select ..  from T where ...
> where T contains a very large number of partitions does not work very well with CombineHiveInputFomat.
> A pool is created per directory, which leads to a high number of mappers.
> In case all partitions share the same operator tree, and the same partition description,
only a single pool should be created.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


Mime
View raw message