[ https://issues.apache.org/jira/browse/HDFS-4606?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13603995#comment-13603995 ]
Sanjay Radia commented on HDFS-4606: ------------------------------------ A long running service such as HBASE region server sometimes relocates; in such a situation it is desirable to relocate replicas for the file blocks such a server reads to the server's location. (See the discussion in HBASE-4755). The alternate idea of co-locating blocks of HBase's "store files" would provide lower latency after failover compared to this proposal; however this proposal is also useful in general, easy to implement and would, I suspect, solve a good part of the region server performance-after-failover issue. > HDFS API to move file replicas to caller's location > --------------------------------------------------- > > Key: HDFS-4606 > URL: https://issues.apache.org/jira/browse/HDFS-4606 > Project: Hadoop HDFS > Issue Type: New Feature > Reporter: Sanjay Radia > Assignee: Sanjay Radia > -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira