> > [I]sn't the AUI interface a half-duplex device?
>
> Yes, AUI is half-duplex; it is designed for interfacing to shared
> broadcast media like 10base5 or 10base2 which are inherently
> half-duplex.
>
> However, I didn't see anyone talking about AUI anywhere, so I'm not
> sure why that's relevant.

Yes, sorry - the reason its relevant is that up until getting the bigmac
card, I've been using AUI-to-Ethernet transeivers with the switch in
question without any problems.


> > Another thing I noticed was that if I changed the
> > "local-mac-address?" OB env.var. to "true", I didn't get that
> > stuttering LED on the switch.
>
> Does the machine in question have multiple interfaces, which are
> connected to different sides of the router, or different ports on the
> same switch, or any such?

It does have multiple interfaces, but none are connected to anything. In a
previous post I mentioned that the TPE port doesn't appear to work on this
machine. Don't know about the AUI ports as I don't have the weird adapter
required to connect it them (some sort of high density thing called a
Mini-26).

> > Using the same unicast MAC address more than once within a given ARP
> domain (hub, vlan on a switch, etc) is a Bad Thing.  If the switch has
> vlans, there is potential for its getting confused by seeing the same
> MAC address in multiple vlans.  Such behaviour is buggy, and most
> switches these days don't suffer from it, but the potential is there
> and depending on how many corners were cut when your switch was made it
> could be relevant.
>
> Of course, this is entirely guessing.  I'm just trying to come up with
> reasons why setting local-mac-address? true might make a difference.
> (This is especially odd because I think NetBSD doesn't pay any
> attention to local-mac-address?.)

Actually, I'm beginning to think that setting local-mac-address?= true has
nothing to do with this card suddenly starting to work. At some point I
unplugged everything (both routers and switch) and upon hooking everything
back up and reseting /etc/mygate back to SMC_Router, the card started
working. Guess maybe it was invariable flushing of the stored routing
tables in the routers and switch, no?

Thanks for all the feedback - you guys are great!

Beaker

--
Overflow on /dev/null, please empty the bit bucket.