Thanks for all the replies. Hope you don't mind being pestered with a couple
of followup questions. Several of the responses pointed out that one needs
to have two IP addresses per system (one each for the ethernet interface and
the FDDI interface).
I'm aware that we need a separate IP address for each <system,interface> pair.
What isn't clear to me is whether the IP addresses for the two systems' FDDI
interfaces need to be on a separate subnet. (One of the replies said that it
would need to be (which is what I would have guessed, but am hoping otherwise),
while the others didn't make any mention of this one way or the other.)
Our ethernet is a subnet of the 128.100 class B network (specifically, we're
assigned 128.100.76). If (as I suspect/fear) the FDDI stuff needs to be on a
different subnet from our ethernet, then it looks like we have only two options:
(1) Partition our existing subnet into two (sub)subnets, using a netmask of
0xffffff80, and assigning 128.100.76.[1-126] to the ethernet connections, and
128.100.76.{129,130} to the two FDDI connections.
(2) Continue using a netmask of 0xffffff00 for our ethernet (using the entire
128.100.76.* range for that network), and get a totally separate class C net
number for the FDDI connection.
Option #1 would be a nuisance, since (for historical reasons) a number of our
systems are already in the upper half of the 128.100.76.* range, and they would
have to be renumbered. And all our existing systems would have to start using
a different netmask.
On the other hand, option #2 would require that we ask the Powers That Be for
a class C network number, which they'd probably give us, but it seems like a
real waste to use up a whole class C net number for a two-machine connection.
I seem to recall reading that there was a proposal (maybe even an RFC) that
a small number of IP network numbers (at least one each in class A, B, and C)
be set aside as "internal use only" networks, i.e. their net numbers could be
used independently by lots of different organizations, without any conflict
(since routes to them wouldn't be advertised outside the organizations using
them). Anyone know whether this was actually implemented?
On the other hand, aside from all of this, the best solution would be if the
FDDI interfaces could be assigned IP addresses in the same range as all our
other IP addresses, using the same netmask. I *think* this isn't permitted,
but it would be nice to get an authoritative answer. (Maybe it's time to go
out and buy a copy of Comer and Stevens!)
If it is allowed, I suspect that it would require a bit of legerdemain with
the routing to get everything to work correctly. So if there's anyone out
there who's pretty networking-savvy, and would be willing to ...
-- let me know whether I am in fact constrained to choosing either option
#1 or option #2 above; and/or ...
-- give me a pointer to more information on the "internal use" IP network
numbers, if such a thing exists; and/or ...
-- if option #3 (assigning the FDDI interfaces IP addresses belonging to
the subnet as our ethernet interfaces) is actually permitted, what needs
to be done (static routing?) to ensure that (a) traffic between the Sables
goes via FDDI, but (b) other traffic (Sable<=>non-Sable) goes via ethernet?
... I'd be vastly appreciative, and forever grateful. (Well, grateful for
at least a few weeks. ;-)
Anyway, thanks again to the people who replied to my first query, namely (in
the order in which the messages arrived) ...
Dave Cherkus <cherkus_at_UniMaster.COM>
John Stoffel <john_at_WPI.EDU>
Selden E Ball Jr <SEB_at_LNS62.LNS.CORNELL.EDU>
Jan.Berger.Henriksen_at_ii.uib.no
"Suzanne.Hassell_at_jhuapl.edu" <sh_at_aurora.jhuapl.edu>
Jon Forrest <forrest_at_postgres.Berkeley.EDU>
fabrice_at_cisk.atmos.Ucla.EDU (Fabrice Cuq)
"Dr. Tom Blinn, 603-881-0646" <tpb_at_zk3.dec.com>
"nagroM W sirdI" <im_at_uvo.dec.com>
Peter Fogarty <syspjf_at_devetir.qld.gov.au>
Henning Fruechtenicht <fruechte_at_rz5.rz.fh-wilhelmshaven.de>
Mark Bartelt 416/978-5619
Canadian Institute for mark_at_cita.toronto.edu
Theoretical Astrophysics mark_at_cita.utoronto.ca
"Sheep not busy being shorn are busy frying" - Dylan, at a NZ lamb barbecue
[ singing "It's all right, ma (I'm only bleating)" ]
Received on Wed Mar 08 1995 - 11:28:48 NZDT