On Wed, 06 Sep 2000, David Brownlee wrote:
> Can you ping -n <ip_of_linux_box> from the NetBSD box?

When I got your note this AM, no, I could not. ping reported that the host was
down as it had been (which it wasn't as the two boxes are about 2.5 feet apart).

To make sure that the ethernet worked well, I turned off packet forwarding and
ipchains on the linux machine and I booted both boxes into MacOS. Each box
was able to see the other. Hardware is OK.

Both boxes were then booted back into their respective unix flavor. Each box
can now ping the other again, but I note now that pings from BSD to Linux take
anywhere from 6-24 msec (I don't remember that much delay prior), but are near
instantaneous (0.5 msec) in the other direction (Linux>BSD) or if I ping the
BSD box from itself. Not sure whether this is something to be concerned about
or not.

Any network activity in or out of the BSD machine generates scads of "mc0:
collision error" msgs on the screen. Dave Huang has told me that this is
normal, but can I suppress them somehow (short of a kernel recompile or upgrade
to 1.5 alpha, which I can do but would rather get current problems under
control first)? The error msgs make telnet etc nearly unreadable.

I'll be trying to set up ip forwarding and masq again this evening to see if
the same problems occur.

Enough rambling for now -- thanks for listening : ). Suggestions gratefully
accepted...

Pat

-- 
Powered by Yellow Dog Linux
--
Pat Plummer.MD
ppat@micron.net
ppat@mac.com