[jira] Commented: (HADOOP-1766) Merging Block and BlockInfo classes on name-node.

["Hadoop QA (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/HADOOP-1766?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12522019 ] 

Hadoop QA commented on HADOOP-1766:
-----------------------------------

+1

http://issues.apache.org/jira/secure/attachment/12364374/blockInfo.patch applied and successfully tested against trunk revision r568809.

Test results:   http://lucene.zones.apache.org:8080/hudson/job/Hadoop-Patch/601/testReport/
Console output: http://lucene.zones.apache.org:8080/hudson/job/Hadoop-Patch/601/console

> Merging Block and BlockInfo classes on name-node.
> -------------------------------------------------
>
>                 Key: HADOOP-1766
>                 URL: https://issues.apache.org/jira/browse/HADOOP-1766
>             Project: Hadoop
>          Issue Type: Improvement
>          Components: dfs
>    Affects Versions: 0.14.0
>            Reporter: Konstantin Shvachko
>            Assignee: Konstantin Shvachko
>             Fix For: 0.15.0
>
>         Attachments: blockInfo.patch
>
>
> In current implementation
> - BlocksMap references BlockInfo
> - INode references Block
> - BlockInfo contains a reference to the corresponding Block
> It would be better to incorporate Block into BlockInfo and reference the latter everywhere in name-node structures.
> This saves 24 bytes per block on a 64-bit jvm as stated in HADOOP-1687 (5).
> I retained the Block class as a structure for external (client and data-node) communication.

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