-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Fri, 4 Oct 2002, Ivan Popov wrote:

> You do not tell us what platform and kernel version you are running,
> but anyway - I would definitely upgrade to the latest stable Coda
> version (5.3.19).

For the record, Debian GNU/Linux with a near-stock 2.4.14 kernel on both
the client and the server.  Minor modifications like ALSA, FreeS/WAN,
etc.

> You may copy the data to the local disk or to some other media (on the
> concerned client), before reinitializing venus.

Took a quick tarball (it's a small filesystem, 222k), then reinitialised
venus.  Definately seems to have fixed the problem.  It's holding in
fully-connected state, without needing strong mode.

Further, although 'cfs fr' still reports 'Invalid argument', it does
actually seem to force syncs.  As does 'cfs wr', so it looks like all is
normal.

It might have something to do with my accidentally downloading a big
Linux kernel tarball into my Coda directory (ooops); at about 140
kbits/sec upload capacity, it may have forced me into weakly-connected
mode.  Why it wasn't able to reintegrate (after I moved that thing out
of there quickly) remains a mystery.

In any case, to my untrained eye, 'unrepaired local subtree(s)' *could*
suggest a venus cache error.  (Of course, I'd be more inclined to
attribute that sort of message to a global/local reintegration conflict,
but there didn't seem to be any.)

Thanks for your reply; glad to have it all working again.  Have upgraded
both the server and all clients to 5.3.19.

-----BEGIN PGP SIGNATURE-----
Version: PGP 6.5.1

iQA/AwUBPZ38zv8wZ8HVeo8aEQIv0QCgr0UqaeMPv95jXGIp00M3A7MH8VUAoNvL
IUGdTWwgVyl1CXSdfqt1qF+C
=65J7
-----END PGP SIGNATURE-----