Nick Gustas skrev:
Kasper Revsbech wrote:
I have attached to windows xp clients by crossed cable one to each if
(vr and vr2)
The fun begins here, when i turn on and off the windows machines a
couple of times one of the can't obtain a IP. It actually brings
down the whole interface. I can't attach another BSD machine and
run dhclient. Even if I set my own ip-addr on the client and starts
to ping I can't reach the other machine. It dosn't even answer on arp
on that if.
This is only solved by rebooting til gateway machine.
I have tried all combinations. like switching cables interfaces and
so on. But I can't relate it to one windows machine, or one interface
on the gateway, and it happens even thought I switch cables.
I don't get any messages in /var/log/daemons from the interface after
it is down but the one still working keeps renewing.
I don't get any warnings in dmesg
I just attach it in the bottom...
I am realy stuck. I have tried to disable pf before and after the
interface has gone down, but still the same. I even tried to take the
take the disk to my soekris 4801 and switched the interface names and
so on, but still the same...
I believe you are hitting a bug with vr, not sure if it's vr in
general or just vr on the 5501. Try running 'ifconfig vr2 down'
'ifconfig vr2 up' and see if that fixes vr2. It may or may not have
been fixed in -current, I haven't been following this issue.
Related thread:
http://kerneltrap.org/mailarchive/openbsd-misc/2007/10/12/336090
I ran into the same problem on a 5501 I was deploying, it could be
triggered by simply unplugging and replugging the patch cable a few
times, whatever vr interface it tried it on would be unusable,
ifconfig down and up would fix it. I only needed two ports and
couldn't risk this happening in the field, so I put a dual port intel
card in to work around the problem.
Exactly...
The behaviour is the same, a de attached cable (Turned of machine) and
then bang....
Anyone aware if this is fixed in current?