Greg Troxel <gdt_at_ir.bbn.com> writes:

> I bring interfaces down (deleting addresses first) before
> suspending, and then bring them up at the new place.  I tend not to
> have both wired and wireless up at the same time.

Hmm.  I've had some trouble with ethernet and suspending.  If I don't
bring the interface down, and then up again, it quietly doesn't work
after the resume.  I now have apmd run suspend and resume scripts that
handle this for me.

> I have also patched dhclient to catch SIGTERM and remove the
> addresses.

I'm doing it slightly differently now -- a combination of apmd and
ifwatchd scripts, and the OMAPI interface to dhclient.  All that's
missing is a good way to give the UTP interface a higher priority than
the wireless one when choosing a default route.

> Do you still have an interface up with an address that doesn't work
> (e.g., address on wired net with cable not plugged in)?

Yes, I have been doing that.  That is, the interface has been left up,
but the address has been removed (set to 0.0.0.0 by dhclient), and the
routes out it flushed.

> If you 'ifconfig down' non-functional interfaces, is venus then able
> to connect to the server?

I will try this on Monday.  Today, I'm at home with a sick child, and
I don't (yet) have wireless access here.  I have changed my setup, and
it works fine here, but I'm looking forward to testing with wireless.

> This is wrong (multiprocessing hazard), but it works all the time
> for me.  netbsd dhclient fix to remove addresses at exit:

Using the OMAPI interface to dhclient to stop it, it does this as is.
It can also be suspended (releasing its leases and unconfiguring its
interfaces), and resumed -- which is what I do at suspend/resume now.

Thanks for your help, Greg!

-tih
-- 
Tom Ivar Helbekkmo, Senior System Administrator, EUnet Norway
www.eunet.no  T: +47-22092958 M: +47-93013940 F: +47-22092901