tajo-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jihoon Son (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (TAJO-464) Rename the name 'partition', actually meaning shuffle to 'shuffle'.
Date Wed, 01 Jan 2014 23:27:50 GMT

    [ https://issues.apache.org/jira/browse/TAJO-464?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13859962#comment-13859962
] 

Jihoon Son commented on TAJO-464:
---------------------------------

Thanks, Hyunsik. This is a definitely necessary work.
However, I'm wondering that 'shuffle' is the proper name in our query execution.  As you know,
the name of 'shuffle' comes from a phase of Hadoop/MapReduce that includes several processes
such as sort and data repartition.
Meanwhile, in our query execution, the intermediate data are just repartitioned between the
execution blocks.
So, in my opinion, 'repartition' is the more proper name than 'shuffle'.

I left some more comments at RB.

> Rename the name 'partition', actually meaning shuffle to 'shuffle'.
> -------------------------------------------------------------------
>
>                 Key: TAJO-464
>                 URL: https://issues.apache.org/jira/browse/TAJO-464
>             Project: Tajo
>          Issue Type: Improvement
>            Reporter: Hyunsik Choi
>            Assignee: Hyunsik Choi
>             Fix For: 0.8-incubating
>
>         Attachments: TAJO-464.patch, TAJO-464_2014-01-01_22:59:40.patch
>
>
> So far, we have used the word 'partition' to indicate shuffle. This name was originated
from repartition in the database studies. However, this name makes us hard to distinguish
the shuffle and table partition. We ned to change such names and prefix into 'shuffle'.



--
This message was sent by Atlassian JIRA
(v6.1.5#6160)

Mime
View raw message