[
https://issues.apache.org/jira/browse/HDFS-5026?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13717913#comment-13717913
]
Andrew Wang commented on HDFS-5026:
-----------------------------------
Eli chatted with me about this one, and we decided it's not really a bug.
getFileStatus/getFileLinkStatus are supposed to be like stat(2), not stat(1).
We only recently got inodes in HDFS, so FileStatus was using this path for a
similar purpose to an inode #. Thus, this path is meant to be opaque.
I suspect there's be code out there making incorrect assumptions about this
path, but these functions themselves are not in error.
> getFileStatus and getFileLinkStatus do not properly handle intermediate
> symlinks in a path
> ------------------------------------------------------------------------------------------
>
> Key: HDFS-5026
> URL: https://issues.apache.org/jira/browse/HDFS-5026
> Project: Hadoop HDFS
> Issue Type: Bug
> Affects Versions: 3.0.0, 2.1.0-beta
> Reporter: Andrew Wang
>
> Imagine a filesystem with the following:
> {noformat}
> /dir
> /dir/file
> /dirLink -> /dir
> {noformat}
> When calling getFileStatus or getFileLinkStatus on {{/dirLink/file}}, the
> returned FileStatus incorrectly has the resolved path of {{/dir/file}}. This
> differs from expectations and the Unix {{stat(1)}} command, which would
> return {{/dirLink/file}}.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
