You are viewing a plain text version of this content. The canonical link for it is here.
Posted to common-issues@hadoop.apache.org by "Andrew Wang (JIRA)" <ji...@apache.org> on 2016/10/07 20:38:20 UTC
[jira] [Commented] (HADOOP-11798) Native raw erasure coder in XOR
codes
[ https://issues.apache.org/jira/browse/HADOOP-11798?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15556179#comment-15556179 ]
Andrew Wang commented on HADOOP-11798:
--------------------------------------
I think this one is pretty important for XOR performance.
[~drankye], related question, do we have an EC policy for XOR? This would be nice for small clusters, since right now the smallest stripe width is (3,2) which requires at least 5 racks. An XOR (2,1) would only require 3.
> Native raw erasure coder in XOR codes
> -------------------------------------
>
> Key: HADOOP-11798
> URL: https://issues.apache.org/jira/browse/HADOOP-11798
> Project: Hadoop Common
> Issue Type: Sub-task
> Components: io
> Reporter: Kai Zheng
> Assignee: Kai Zheng
> Labels: hdfs-ec-3.0-must-do
> Fix For: HDFS-7285
>
> Attachments: HADOOP-11798-v1.patch
>
>
> Raw XOR coder is utilized in Reed-Solomon erasure coder in an optimization to recover only one erased block which is in most often case. It can also be used in HitchHiker coder. Therefore a native implementation of it would be deserved for performance gain.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: common-issues-unsubscribe@hadoop.apache.org
For additional commands, e-mail: common-issues-help@hadoop.apache.org