hadoop-pig-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Olga Natkovich (JIRA)" <j...@apache.org>
Subject [jira] Commented: (PIG-545) PERFORMANCE: Sampler for order bys does not produce a good distribution
Date Tue, 17 Feb 2009 23:30:59 GMT

    [ https://issues.apache.org/jira/browse/PIG-545?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12674393#action_12674393

Olga Natkovich commented on PIG-545:

The patch looks good. A couple of comments/questions:

(1) One comment that I have is that MalFormedProbVecException does not follow new error handling
practices. It is not derived from PigException and it does not have a way to pass error code
to it.

(2) My understanding is that weighted partitioner would actually add overhead to well randomly
distributed data. I am fine saying that this case is not very common or that the overhead
is not great enough to worry about but I wanted to check what other committers think.

(3) WeightedRangePartitioner.java throws RuntimeException - this is not consistent with our
new error handling.

> PERFORMANCE: Sampler for order bys does not produce a good distribution
> -----------------------------------------------------------------------
>                 Key: PIG-545
>                 URL: https://issues.apache.org/jira/browse/PIG-545
>             Project: Pig
>          Issue Type: Bug
>          Components: impl
>    Affects Versions: types_branch
>            Reporter: Alan Gates
>            Assignee: Pradeep Kamath
>             Fix For: types_branch
>         Attachments: PIG-545-v3.patch, WRP.patch, WRP1.patch
> In running tests on actual data, I've noticed that the final reduce of an order by has
skewed partitions.  Some reduces finish in a few seconds while some run for 20 minutes.  Getting
a better distribution should lead to much better performance for order by.

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

View raw message