hadoop-hive-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Namit Jain (JIRA)" <j...@apache.org>
Subject [jira] Commented: (HIVE-1509) Monitor the working set of the number of files
Date Wed, 04 Aug 2010 00:32:21 GMT

    [ https://issues.apache.org/jira/browse/HIVE-1509?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12895106#action_12895106

Namit Jain commented on HIVE-1509:

With dynamic partitions, the number of intermediate files can grow very fast.

For example, consider a query with 10,000 mappers and 100 files per mapper - it can create
up to 1 million files before merging them at the end.
The cluster may be down by the time the query finishes.

It is a good idea to track the number of files through a counter, and kill the query if the
number exceeds a given threshold

> Monitor the working set of the number of files 
> -----------------------------------------------
>                 Key: HIVE-1509
>                 URL: https://issues.apache.org/jira/browse/HIVE-1509
>             Project: Hadoop Hive
>          Issue Type: Bug
>          Components: Query Processor
>    Affects Versions: 0.7.0
>            Reporter: Namit Jain
>            Assignee: Ning Zhang

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

View raw message