On Jul 19, 2007, at 11:23 AM, Albert Chin wrote:

> We have a X4100 running b66 exporting ZFS filesystems via NFS to
> several HP-UX clients. Due to bug 6538387,
> http://bugs.opensolaris.org/bugdatabase/view_bug.do?bug_id=6538387, we
> must mount the ZFS filesystems from our X4100 server as NFSv2.

   Just to be clear, that bug was filed against the Solaris
   NFS server; however, the evaluation cleared our server
   and implicated the HP-UX client.  The bug was closed back
   in April.  I've attached the bug evaluation.

> However, this causes problems with getconf(1) on this system:
>   hp-ux# mount -o vers=2 b66:/opt/tww /mnt
>   hp-ux# getconf NAME_MAX /mnt
>   getconf NAME_MAX /mnt: Operation not supported
>
>   hp-ux# mount -o vers=3 b66:/opt/tww /mnt
>   hp-ux# getconf NAME_MAX /mnt
>   255
>

   NFS2 has no protocol support for querying the server's
   pathconf capabilities.  The NFS3 protocol uses the FSINFO
   procedure to convey pathconf info between the server
   and client.  The NFS4 protocol defines attributes to
   query for support and existence of named attrs.

   On a Solaris client, getconf NAME_MAX over a NFS2 mount simply
   prints "undefined".  Without protocol support, it isn't possible
   to query NAME_MAX in a way that will interoperate with NFS2
   servers from different vendors.

> We can "zfs set xattr=off" on the ZFS filesystems as, according to
> zfs(1m):
>   xattr=on | off
>       Controls whether extended  attributes  are  enabled  for
>       this file system. The default value is "on".
>
> However, this isn't enough. Is it possible to disable the extra
> attribute bits used by NFS when NFSv3 is being used by the client? We
> don't need any extra attribute bits more than the standard UNIX
> permissions.
>

This won't fix the getconf NAME_MAX issue over NFS2, but since
you asked...

I can't speak for the HP client, but in Solaris, support for named
attrs can be disabled using the -o noxattr mount option.  This
should work for all native Solaris file systems (NFS, ZFS, UFS, tmpfs,
etc).  Setting the ZFS xattr attribute as you did would work too.

Just FYI, but named attr support is not related to getconf NAME_MAX.
NAME_MAX describes the max length of a single path component.  The
Solaris pathconf() cmds related to named attrs are:

     _PC_XATTR_ENABLED:  does the file system support named attrs?
     _PC_XATTR_EXISTS:   do any named attrs exist for this file/dir?

I don't know of any "standardized" pathconf cmds to query
named attr support, so they will most likely be different
on HP-UX.

Jeff


-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: hpbug.txt
URL: 
<http://mail.opensolaris.org/pipermail/nfs-discuss/attachments/20070719/b1e7fe2f/attachment.txt>

Reply via email to