hadoop-common-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Uma Maheswara Rao G (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HADOOP-11569) Provide Merge API for MapFile to merge multiple similar MapFiles to one MapFile
Date Wed, 11 Feb 2015 13:41:12 GMT

    [ https://issues.apache.org/jira/browse/HADOOP-11569?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14316201#comment-14316201

Uma Maheswara Rao G commented on HADOOP-11569:

Thanks Vinay for the patch.
Here is the feedback on the patch.
// Read first key/values from all inputs
It should say "first key/value"

* 4. Read the next key/value of selected input <br>
How about like "Read the next key/value of selected inputReader from which we read least key"

One question: From the merger, open is private and merger is closeable. When open if private,
what is the real use of public close method? Infact close call was already handled internally
from merge method itself. So, could please explain if any other reasons which I might missed.
Merge logic seems fine to me.

> Provide Merge API for MapFile to merge multiple similar MapFiles to one MapFile
> -------------------------------------------------------------------------------
>                 Key: HADOOP-11569
>                 URL: https://issues.apache.org/jira/browse/HADOOP-11569
>             Project: Hadoop Common
>          Issue Type: Improvement
>            Reporter: Vinayakumar B
>            Assignee: Vinayakumar B
>         Attachments: HADOOP-11569-001.patch, HADOOP-11569-002.patch, HADOOP-11569-003.patch
> If there are multiple similar MapFiles of the same keyClass and value classes, then these
can be merged together to One MapFile to allow search easier.
> Provide an API  similar to {{SequenceFile#merge()}}.
> Merging will be easy with the fact that MapFiles are already sorted.

This message was sent by Atlassian JIRA

View raw message