You are viewing a plain text version of this content. The canonical link for it is here.
Posted to hdfs-dev@hadoop.apache.org by "Chris Nauroth (JIRA)" <ji...@apache.org> on 2014/03/31 22:04:15 UTC
[jira] [Resolved] (HDFS-241) ch{mod,own,grp} -R to do recursion at
the name node
[ https://issues.apache.org/jira/browse/HDFS-241?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Chris Nauroth resolved HDFS-241.
--------------------------------
Resolution: Won't Fix
I'm going to close this. The idea of doing recursive operations on the server side has come up in discussion on a few other things recently. The consensus is that we're actually reliant on the client side to drive tree traversal as a form of RPC throttling. (Server-side tree traversal would cause an individual client to consume an RPC handler thread in the NameNode for a longer period of time.) There is also the issue of cross-filesystem symlinks, which require resolution on the client side in case there is a client-side mount table configured or a symlink pointing to a different cluster.
> ch{mod,own,grp} -R to do recursion at the name node
> ---------------------------------------------------
>
> Key: HDFS-241
> URL: https://issues.apache.org/jira/browse/HDFS-241
> Project: Hadoop HDFS
> Issue Type: New Feature
> Reporter: Robert Chansler
>
> Performance. No need to maintain {{distch}}.
--
This message was sent by Atlassian JIRA
(v6.2#6252)