[jira] [Commented] (HADOOP-11921) Enhance tests for erasure coders

["Uma Maheswara Rao G (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/HADOOP-11921?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14530365#comment-14530365 ] 

Uma Maheswara Rao G commented on HADOOP-11921:
----------------------------------------------

Hi Kai,

some nit:

{code}
 // Erase the copied sources
    ECChunk[] erasedChunks = eraseAndReturnChunks(clonedDataChunks);
{code}
Actually returned chunks are not erased chunks right? they are cloned input chunks instead. Am I right? So, need to change variable name to backupChunks?
May be for better readbility, clone them separately and just do erseChunks in eraseAndReturnChunks API?

Other than this changes looks good to me.

> Enhance tests for erasure coders
> --------------------------------
>
>                 Key: HADOOP-11921
>                 URL: https://issues.apache.org/jira/browse/HADOOP-11921
>             Project: Hadoop Common
>          Issue Type: Sub-task
>          Components: io
>            Reporter: Kai Zheng
>            Assignee: Kai Zheng
>         Attachments: HADOOP-11921-v1.patch
>
>
> While working on native coders, it was found better to enhance the tests for erasure coders to:
> * Test if erasure coders can be repeatedly reused;
> * Test if erasure coders can be called with two buffer types (bytes array version and direct bytebuffer version).



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)