[
https://issues.apache.org/jira/browse/HADOOP-8299?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13258597#comment-13258597
]
Eli Collins commented on HADOOP-8299:
-------------------------------------
Here's the exception you get btw if you try to create a slash mount entry. We
should make this a more friendly error message in the mean time.
{noformat}
java.lang.ArrayIndexOutOfBoundsException: 1
at org.apache.hadoop.fs.viewfs.InodeTree.createLink(InodeTree.java:237)
at org.apache.hadoop.fs.viewfs.InodeTree.<init>(InodeTree.java:334)
at org.apache.hadoop.fs.viewfs.ViewFileSystem$1.<init>(ViewFileSystem.java:165)
at
org.apache.hadoop.fs.viewfs.ViewFileSystem.initialize(ViewFileSystem.java:165)
at org.apache.hadoop.fs.FileSystem.createFileSystem(FileSystem.java:2041)
at org.apache.hadoop.fs.FileSystem.access$200(FileSystem.java:78)
at org.apache.hadoop.fs.FileSystem$Cache.getInternal(FileSystem.java:2075)
at org.apache.hadoop.fs.FileSystem$Cache.get(FileSystem.java:2057)
at org.apache.hadoop.fs.FileSystem.get(FileSystem.java:284)
at org.apache.hadoop.fs.FileSystem.get(FileSystem.java:151)
{noformat}
> ViewFs doesn't work with a slash mount point
> --------------------------------------------
>
> Key: HADOOP-8299
> URL: https://issues.apache.org/jira/browse/HADOOP-8299
> Project: Hadoop Common
> Issue Type: Bug
> Components: fs
> Affects Versions: 2.0.0
> Reporter: Eli Collins
>
> We currently assume [a typical viewfs client
> configuration|https://issues.apache.org/jira/secure/attachment/12507504/viewfs_TypicalMountTable.png]
> is a set of non-overlapping mounts. This means every time you want to add a
> new top-level directory you need to update the client-side mountable config.
> If users could specify a slash mount, and then add additional mounts as
> necessary they could add a new top-level directory without updating all
> client configs (as long as the new top-level directory was being created on
> the NN the slash mount points to). This could be achieved by HADOOP-8298
> (merge mounts, since we're effectively merging all new mount points with
> slash) or having the notion of a "default NN" for a mount table.
--
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
