hadoop-hdfs-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Todd Lipcon (Commented) (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HDFS-1973) HA: HDFS clients must handle namenode failover and switch over to the new active namenode.
Date Mon, 03 Oct 2011 20:24:36 GMT

    [ https://issues.apache.org/jira/browse/HDFS-1973?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13119567#comment-13119567
] 

Todd Lipcon commented on HDFS-1973:
-----------------------------------

Aaron and I just chatted about this a bit. Here's a summary of what we discussed:
- the if condition in {{performFailover}} was somewhat confusing to me as to its purpose.
Aaron explained that its purpose is to avoid the case where multiple outstanding RPC calls
fail, and then they all call performFailover at the same time. If there were an even number
of such calls, and you didn't do any such checks for "already failed over", then you'd have
a case where you failover twice and end up back at the original proxy object.
- we decided that, rather than try to handle this situation in the FailoverProvider itself,
it would be better to do this at the caller. Otherwise, each failover provider implementation
will have to have this same concern.

So, Aaron is going to update the patch to include a safeguard at the call site of {{performFailver}}
which checks that, before calling performFailover, another thread hasn't already failed over
to a new proxy object.
                
> HA: HDFS clients must handle namenode failover and switch over to the new active namenode.
> ------------------------------------------------------------------------------------------
>
>                 Key: HDFS-1973
>                 URL: https://issues.apache.org/jira/browse/HDFS-1973
>             Project: Hadoop HDFS
>          Issue Type: Sub-task
>            Reporter: Suresh Srinivas
>            Assignee: Aaron T. Myers
>         Attachments: HDFS-1973-HDFS-1623.patch, HDFS-1973-HDFS-1623.patch, HDFS-1973-HDFS-1623.patch,
hdfs-1973.0.patch
>
>
> During failover, a client must detect the current active namenode failure and switch
over to the new active namenode. The switch over might make use of IP failover or some thing
more elaborate such as zookeeper to discover the new active.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

Mime
View raw message