[
https://issues.apache.org/jira/browse/HADOOP-11588?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15416474#comment-15416474
]
Kai Zheng commented on HADOOP-11588:
------------------------------------
Thanks [~lirui] for the work! The patch LGTM. Only some minors:
1. Do you have some rational not to place the micro benchmark tool in the test
scope?
2. In the benchmark tool codes, could we refactor some bit to avoid typos like
'client'? It's a cmd tool without connecting any server.
3. In the benchmark test codes: note {{public void testISALCOder()}} a typo;
could we use enum names to mention the coder index?
> Benchmark framework and test for erasure coders
> -----------------------------------------------
>
> Key: HADOOP-11588
> URL: https://issues.apache.org/jira/browse/HADOOP-11588
> Project: Hadoop Common
> Issue Type: Sub-task
> Components: io
> Reporter: Kai Zheng
> Assignee: Rui Li
> Attachments: HADOOP-11588-HDFS-7285.2.patch, HADOOP-11588-v1.patch,
> HADOOP-11588.3.patch, HADOOP-11588.4.patch, HADOOP-11588.5.patch,
> HADOOP-11588.6.patch, HADOOP-11588.7.patch
>
>
> Given more than one erasure coders are implemented for a code scheme, we need
> benchmark and test to help evaluate which one outperforms in certain
> environment. This is to implement the benchmark framework.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
