hadoop-mapreduce-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Arun C Murthy (JIRA)" <j...@apache.org>
Subject [jira] Commented: (MAPREDUCE-326) The lowest level map-reduce APIs should be byte oriented
Date Tue, 16 Feb 2010 21:05:28 GMT

    [ https://issues.apache.org/jira/browse/MAPREDUCE-326?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12834468#action_12834468
] 

Arun C Murthy commented on MAPREDUCE-326:
-----------------------------------------

bq. Doug> New public APIs and new code can be a burden as they must be supported, so they
must provide some clear benefits.

Exactly. 

This argument isn't about priorities, it is about an incremental change versus a whole new
set of apis.

As someone who both worked on all parts of the framework and as some who maintains lots of
it I do *not* see any clear benefit of adding a _whole_ set of _new apis_ we need to support
and maintain since the incremental api change that Owen proposes gives all of the control
necessary to accomplish _all_ of the stated goals being proposed by Tom at a *significantly*
smaller cost of maintenance. 

I haven't heard a single argument about the drawbacks of the incremental change, and why they
won't suffice - could you please enlighten us?

> The lowest level map-reduce APIs should be byte oriented
> --------------------------------------------------------
>
>                 Key: MAPREDUCE-326
>                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-326
>             Project: Hadoop Map/Reduce
>          Issue Type: Improvement
>            Reporter: eric baldeschwieler
>         Attachments: MAPREDUCE-326-api.patch, MAPREDUCE-326.pdf
>
>
> As discussed here:
> https://issues.apache.org/jira/browse/HADOOP-1986#action_12551237
> The templates, serializers and other complexities that allow map-reduce to use arbitrary
types complicate the design and lead to lots of object creates and other overhead that a byte
oriented design would not suffer.  I believe the lowest level implementation of hadoop map-reduce
should have byte string oriented APIs (for keys and values).  This API would be more performant,
simpler and more easily cross language.
> The existing API could be maintained as a thin layer on top of the leaner API.

-- 
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