[
https://issues.apache.org/jira/browse/HDFS-6961?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Jing Zhao updated HDFS-6961:
----------------------------
Attachment: HDFS-6961.000.patch
Initial patch. The patch also contains a unit test where the original mechanism
fails to return enough datanodes.
> BlockPlacementPolicy#chooseTarget should check each valid storage type in
> each choosing round
> ---------------------------------------------------------------------------------------------
>
> Key: HDFS-6961
> URL: https://issues.apache.org/jira/browse/HDFS-6961
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: balancer, namenode
> Reporter: Jing Zhao
> Assignee: Jing Zhao
> Attachments: HDFS-6961.000.patch
>
>
> Currently in each choosing round of BlockPlacementPolicy#chooseTarget (e.g.,
> round 1 for choosing local storage, round 2 for remote rack, etc.), we only
> take into account the first storage type in the storage type list. This may
> fail to choose the local machine or sometimes may even fail to choose enough
> nodes.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
