Re: Linux NFS server problem

Zygo Blaxell (uixjjji1@umail.furryterror.org)
18 Feb 1999 14:01:10 -0500

In article <199902180252.SAA02909@Harpo.ixlabs.com>,
G. Allen Morris III <gam3@harpo.ixlabs.com> wrote:
>H.J., I just don't understand the problem that you are trying to solve.
>I 2.3.3 of RFC 1094 it stats:
>
> The file handle can contain whatever information the
> server needs to distinguish an individual file.
>
>WHATEVER is pretty strong language!
>I can see cases where a system might put a timestamp in every
>filehandle or a counter. If solaris is complaining because the
>NFS filehandle is different between two calls, then it is broken.
>
>APPENDIX
>
>ftp://ftp.isi.edu/in-notes/rfc1094.txt
>
>2.3.3. fhandle
>
> typedef opaque fhandle[FHSIZE];
>
> The "fhandle" is the file handle passed between the server and the
> client. All file operations are done using file handles to refer
> to a file or directory. The file handle can contain whatever
> information the server needs to distinguish an individual file.

Last time I checked, the filehandle for any particular file had to be
_constant_ from a client's point of view, even if the server reboots.
You could for example hash and encrypt the device and inode number of
each file and the IP address of the client with a secret key and use
that as an NFS filehandle. However, you can't use a timestamp in there
unless you are willing to accept all possible values of the timestamp
as aliases for the same file (and you're likely to cause some severe
caching inconsistency on the client if you do this anyway).

-- 
Zygo Blaxell, Linux Engineer, Corel Corporation, zygob@corel.ca (work),
zblaxell@furryterror.org (play).  It's my opinion, I tell you! Mine! All MINE!
Size of 'diff -Nurw [...] winehq corel' as of Thu Feb 18 13:14:00 EST 1999
Lines/files:  In 3905 / 14, Out 30668 / 391, Both 34339 / 394


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/