lkml.org 
[lkml]   [1999]   [Feb]   [8]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
SubjectRe: Kernel interface changes (was Re: cdrecord problems on
Date
From
In message <19990207193840.B3051@bug.ucw.cz>, Pavel Machek writes:
+-----
| > i know Linus was just picking on AFS as an example, but AFS is one of
| > those weird support issues -- not many people use it, but those who do,
| > depend on it for everything.
| Ook, so you depend on AFS. Do you need speed? If you can afford to
| loose a _lot_ of speed, you can develop userland nfs daemon that will
| re-export AFS. That way, you'll have portable & kernel-change-proof
| solution.
+--->8

Actually, that's a standard part of AFS clients on most platforms (but not
Linux as yet... sigh). Problem is, AFS implements token-based permissions,
and NFS doesn't deal with those very well. Otherwise, Linux's user-space
nfsd is perfectly happy to re-export /afs as an NFS mount.

The "AFS to NFS translator", which is part of full AFS client distributions
(hence not yet available for Linux), re-exports AFS as NFS and also manages
tokens. The "knfs" program (which *is* available for Linux, but I'm not sure
how well tested it is) can be used on NFS clients that mount the exported AFS
from the translator host to manage AFS tokens remotely.

The problem I've found is that the NFS translator seems not to work well on
Suns for some reason --- at least, not in the sun4_551 version. Possibly 2.6
does better, but we're stuck on 2.5.1 until certain programs get ported (and
no, I have even less say in switching away from those programs than I have
about AFS). A quick test with an AIX host suggests that it works better
there, but the AIX boxes we have are generally underpowered for the task and
moderately temperamental.

I've thought about the idea of using a (modified, if necessary) Samba on an
exporter host and smbfs on the Linux side, since Samba has a better concept
of authentication in the AFS fashion; problem there being that I can't control
whether the custom kernels that are my biggest problem get built with smbfs
support. (---My life would be much easier if ECE weren't mostly composed of
"appliance operators"... except for the ones who need to hack kernels, but
are better versed in the MIT exokernel than the Linux kernel.)

--
brandon s. allbery [os/2][linux][solaris][japh] allbery@kf8nh.apk.net
system administrator [WAY too many hats] allbery@ece.cmu.edu
carnegie mellon / electrical and computer engineering KF8NH
We are Linux. Resistance is an indication that you missed the point.

-
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/

\
 
 \ /
  Last update: 2005-03-22 13:50    [from the cache]
©2003-2011 Jasper Spaans