hadoop-common-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Ankur (JIRA)" <j...@apache.org>
Subject [jira] Updated: (HADOOP-3669) Aggregate Framework should allow usage of MultipleOutputFormat
Date Mon, 07 Jul 2008 11:05:31 GMT

     [ https://issues.apache.org/jira/browse/HADOOP-3669?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel

Ankur updated HADOOP-3669:

    Assignee: Ankur
      Status: Patch Available  (was: Open)

Here is a patch that allows the specification of the output format on the command-line when
using ValueAggregator framework.
I don't think a unit test case is required for this.

> Aggregate Framework should allow usage of MultipleOutputFormat
> --------------------------------------------------------------
>                 Key: HADOOP-3669
>                 URL: https://issues.apache.org/jira/browse/HADOOP-3669
>             Project: Hadoop Core
>          Issue Type: Improvement
>    Affects Versions: 0.17.0
>            Reporter: Ankur
>            Assignee: Ankur
>         Attachments: HADOOP-3669_v1.patch
> Currently the output format is hard-coded to be TextOutputFormat in ValueAggregatorJob
responsible for running aggregation jobs using user-defined value aggregator descriptors.
This prevents the application writer from specifying an alternate output format.
> A good use case from an application's perspective is to have a sub-type of MultipleOutputFormat
set as output format which takes care of redirecting (key, value)
> to different files based on type information encoded in them.
> Applications can extend MultipleTextOutputFormat and define there own multiple output
format but they still can't hook it into value aggregator framework.

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

View raw message