You are viewing a plain text version of this content. The canonical link for it is here.
Posted to hdfs-issues@hadoop.apache.org by "Justin Joseph (JIRA)" <ji...@apache.org> on 2011/07/19 14:07:58 UTC
[jira] [Created] (HDFS-2165) The Intelligent Clients should switch
dynamically according to the Namenode switchover & the operations that fail
during the switching should be retried on the newly Active Namenode.
The Intelligent Clients should switch dynamically according to the Namenode switchover & the operations that fail during the switching should be retried on the newly Active Namenode.
--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: HDFS-2165
URL: https://issues.apache.org/jira/browse/HDFS-2165
Project: Hadoop HDFS
Issue Type: Sub-task
Reporter: Justin Joseph
Here we consider DataNode and DFSClient as clients to the Namenode. When Active Namenode goes down and Standby takes over the role, the clients should understand it and communicate to the newly Active Namenode from that point of time.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HDFS-2165) The Intelligent Clients should
switch dynamically according to the Namenode switchover & the operations
that fail during the switching should be retried on the newly Active
Namenode.
Posted by "Suresh Srinivas (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HDFS-2165?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13067820#comment-13067820 ]
Suresh Srinivas commented on HDFS-2165:
---------------------------------------
Justin, have you looked at subtasks under HDFS-1623. Could we avoid creating duplicate jiras?
> The Intelligent Clients should switch dynamically according to the Namenode switchover & the operations that fail during the switching should be retried on the newly Active Namenode.
> --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
>
> Key: HDFS-2165
> URL: https://issues.apache.org/jira/browse/HDFS-2165
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: data-node, hdfs client, name-node
> Reporter: Justin Joseph
>
> Here we consider DataNode and DFSClient as clients to the Namenode. When Active Namenode goes down and Standby takes over the role, the clients should understand it and communicate to the newly Active Namenode from that point of time.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
[jira] [Commented] (HDFS-2165) The Intelligent Clients should
switch dynamically according to the Namenode switchover & the operations
that fail during the switching should be retried on the newly Active
Namenode.
Posted by "Justin Joseph (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HDFS-2165?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13067649#comment-13067649 ]
Justin Joseph commented on HDFS-2165:
-------------------------------------
The high level approach is mentioned below.
a) The client will have a list of IP addresses of the Active and Standby, which it uses for the connection string.
b) For connecting to Namenode for the first time, the client will try to ping both the Namenodes. Whoever is reachable, client will treat him as the Active Namenode.
c) The information on which Namenode is Active is cached. For further client requests, DFSClient will route them to the Namenode whose URL is cached. This avoids the overhead of finding the Namenode always through trial and error.
d) When the client detects Active Namenode failure, it will start the switching procedure. Client keeps trying to ping both the Namenodes configured, until it can successfully connect to any one of them.
> The Intelligent Clients should switch dynamically according to the Namenode switchover & the operations that fail during the switching should be retried on the newly Active Namenode.
> --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
>
> Key: HDFS-2165
> URL: https://issues.apache.org/jira/browse/HDFS-2165
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: data-node, hdfs client, name-node
> Reporter: Justin Joseph
>
> Here we consider DataNode and DFSClient as clients to the Namenode. When Active Namenode goes down and Standby takes over the role, the clients should understand it and communicate to the newly Active Namenode from that point of time.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira