[
https://issues.apache.org/jira/browse/HDFS-6458?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14011438#comment-14011438
]
Brandon Li commented on HDFS-6458:
----------------------------------
Most likely, the NFS client attribute cache is massed up by the empty attribute
return in some NFS response along with the error status.
> NFS: stale NFS file handle Error for previous mount point
> ---------------------------------------------------------
>
> Key: HDFS-6458
> URL: https://issues.apache.org/jira/browse/HDFS-6458
> Project: Hadoop HDFS
> Issue Type: Bug
> Components: nfs
> Affects Versions: 2.2.0
> Reporter: Yesha Vora
>
> Steps to reproduce:
> 1) Set dfs.nfs.exports.allowed.hosts = <Gateway> rw
> 2) mount nfs on <Gateway>
> 3) Set dfs.nfs.exports.allowed.hosts = <Datanode> rw
> 4) mount nfs on <Datanode>
> Try to access NFS mount point at Gateway. Can't access mount point from
> Gateway.
> {noformat}
> bash: ls /tmp/tmp_mnt
> ls: cannot access /tmp/tmp_mnt: Stale NFS file handle
> {noformat}
> Expected: Mount_point from previous config should be accessible if it is not
> unmounted before config change.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
