hadoop-hdfs-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Dian Fu (JIRA)" <j...@apache.org>
Subject [jira] [Created] (HDFS-6467) Format mirror cluster with primary cluster's NamespaceInfo in the runtime
Date Thu, 29 May 2014 09:16:02 GMT
Dian Fu created HDFS-6467:
-----------------------------

             Summary: Format mirror cluster with primary cluster's NamespaceInfo in the runtime
                 Key: HDFS-6467
                 URL: https://issues.apache.org/jira/browse/HDFS-6467
             Project: Hadoop HDFS
          Issue Type: Sub-task
    Affects Versions: 3.0.0
            Reporter: Dian Fu
             Fix For: 3.0.0


When a DFSClient write data to datanodes, it will specify a block pool id which is used by
datanodes to distinguish the federation this
data belongs to. We know that block pool id is generated by namenode when doing format, so
different namenodes will have different block
pool id.
In the sync mode of DR, a DFSClient in the primary cluster will write data to Datanodes both
in primary cluster and mirror cluster. So we
need guarantee that the block pool id of primary cluster and mirror cluster are the same.



--
This message was sent by Atlassian JIRA
(v6.2#6252)

Mime
View raw message