🚀 go-pugleaf

I've got a bit of a problem here.

I'm trying to netboot one NeXT off another NeXT, which is easy enough to do.
The catch, though, is the netboot server is in a different subnet, going
through two routers on an FDDI ring.

The boot goes fine -- /private/tftpboot/boot is loaded, then
/private/tftpboot/mach, then it goes around, says the primary ethernet
interface is en0 [some not quite right number -- where DOES it get that,
anyway?] then it gets its hostname, says whoami: no domain name, and panics
because it can't mount root.

The netboot server is exporting root to anyone who cares to mount it (it's on
an isolated network, so security is not a big concern).  If I boot off SCSI
disk, and mount the netboot server's root as, say, /tmp/otherroot, it works
fine.

Netbooting another NeXT on the same subnet as the netboot server works fine,
and both are configured.  I know the bootparams are set up right (well, I'm
reasonably sure).  I know the ethernet address is entered right, as
-AUTOMATIC- works for hostname and IP address on normal booting (btw, if you
wish to do that through a gateway, you have to move the IP Routing code from
/etc/rc to /etc/rc.boot, to right before it tries to do hostname -AUTOMATIC-).

But it can't mount the root directory and dies.  I know, via router
debugging, that it sends out a sunrpc call to the correct host, presumably to
to the NFS mount.  It does not look like the call's ever making it to the
netboot server, though, by debugging on the other router.

But why, then, does it work when the machine boots up normally?  What's
different? 

And where the heck does the first, seemingly bogus, primary ethernet address
come from, anyway?

Any help would be greatly appreciated.
------
Mike Matthews, matthews@oberon.umd.edu (NeXTmail accepted)
------
When in doubt, do what the President does -- guess.