[
https://issues.apache.org/jira/browse/HDFS-7891?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14366605#comment-14366605
]
Walter Su commented on HDFS-7891:
---------------------------------
Thanks [~zhz], the approach is much simple. Could you help me revert
HADOOP-11676? I'll close that and rewrite the patch in HDFS-7891.
This is What I thought:
1. If I decide to use sorted rack method. I have to optimized rack accessing in
{{NetworkTopology.countNumOfAvailableNodes()}} and
{{NetworkTopology.getNode()}} , and Add HADOOP-11676. I'm not very comfortable
with that.
2. I think random policy with {{maxNodesPerRack}} works well for EC. If the
cluster is small and the burden on Namenode is small, the worst case doesn't
look so bad. If the cluster is midium/large, the policy won't hit worst case.
You can check {{testresult.txt}}
> A block placement policy with best fault tolerance
> --------------------------------------------------
>
> Key: HDFS-7891
> URL: https://issues.apache.org/jira/browse/HDFS-7891
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Reporter: Walter Su
> Assignee: Walter Su
> Attachments: HDFS-7891.002.patch, HDFS-7891.patch,
> PlacementPolicyBenchmark.txt, testresult.txt
>
>
> a block placement policy tries its best to place replicas to most racks.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
