[
https://issues.apache.org/jira/browse/HDFS-2839?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13202692#comment-13202692
]
Jitendra Nath Pandey commented on HDFS-2839:
--------------------------------------------
There could be ways around it for example we can be creative with the format
of nameservice id, something that encodes the hostnames of primary and standby
in the nameservice id. I am not saying it is the best idea, but I am reluctant
to dismiss this issue as too hard to even think about. It is similar to viewfs
issue, but viewfs use cases do require the viewfs authorities to be agnostic to
underlying nameservices. For HA this requirement can be relaxed.
> Nameservice id in file uri could cause issues
> ---------------------------------------------
>
> Key: HDFS-2839
> URL: https://issues.apache.org/jira/browse/HDFS-2839
> Project: Hadoop HDFS
> Issue Type: Sub-task
> Components: ha, hdfs client, name-node
> Affects Versions: HA branch (HDFS-1623)
> Reporter: Jitendra Nath Pandey
> Assignee: Jitendra Nath Pandey
>
> There might be two issues with having a nameservice id in fully qualified
> paths of the files.
> 1) Some applications might be storing the fully qualified paths. They will
> have hostnames from pre-ha installations, and they might break after the
> failover.
> 2) The fully qualified path from an ha cluster, won't be usable from a
> different cluster that doesn't know about that particular namespace id.
--
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
