[plug] Weird network fallover?

[Andrew Furey]

Hi all,

I have a rackmount server running woody, with a LiteOn
(onboard, I suspect) tulip-driver network card. This
machine also has a VMware install of Win98, which we
use for the various Windows requirements (which are
few, fortunately).

The VMware is set up in "host-only" mode, so the Win98
has its own IP address on the same subnet as the
server (as opposed to being masqueraded through it).
We access the Win98 using VNC.

This has worked fine for years (although they were
reinstalled a few months ago), but over the last few
weeks we've been having an increasingly-more-frequent
(and more-annoying) problem. At seemingly random times
(but possibly related to actual usage, since we don't
notice it otherwise) the network will suddenly stop.

The interface is still up with correct settings, the
lights are on, mii-tool reports a connection, but
nothing can ping through to/from either IP. There's no
error messages of any sort when it happens, it just
dies.

The weird thing is, I can simply restart the
networking system (/etc/init.d/networking restart) and
it will be fine (without even reloading the driver).
Until next time, which these days might be <30s...

Has anyone seen anything remotely like this? Any
suggestions (apart from a watchdog daemon / cronjob)?

Thanks,
Andrew

=====
Linux supports the notion of a command line or a shell for the same
reason that only children read books with only pictures in them.
Language, be it English or something else, is the only tool flexible
enough to accomplish a sufficiently broad range of tasks.
                          -- Bill Garrett

Find local movie times and trailers on Yahoo! Movies.
http://au.movies.yahoo.com