Greg Wilson-Lindberg wrote:
Rick Jones wrote:
...
Is there a specific reason you cannot further subnet the 10.0.0.0/8
network for your HDLC channels and eschew the NAT? Then at least
ignoring the issue of "routing" broadcast datagrams (these are IP
broadcasts and not simply ethernet broadcasts right?) reaching-out and
touching the HDLC devices from the WAN would seem to be more
straightforward.
rick jones
visions of many static routing table entries floating through his head
The 10.0.0.0/8 network is an existing network that has 100's of
computers on it.
Not knowing (perhaps I just missed it) "the rest of the story" (a la Paul
Harvey), simply 100's of computers doesn't sound like all that much for a
renumbering - should it come to that.
Are the IPs of the "HDLC machines" similarly constrained?
We are planning on using some 10.x.0.0/16 addresses
to NETMAP to the 172.16.0.0/16 addresses in some limited situations, but
this doesn't solve the problem of routing to the correct HDLC channel to
get to the final destination.
I have not fully parsed the manpage for the ip command, but it does have tidbits
which suggest being able to specify interfaces by name when manipulating routing
tables. If that is correct you should be able to add specific host routes
pointing-out specific interface names if you are indeed forced to put assign the
same IP address to each of the HCLC interfaces on the "gateway."
Earlier you mentioned broadcasts that needed to reach through and touch the HDLC
systems - did you actually mean IP multicast?
rick jones
--
To unsubscribe from this list: send the line "unsubscribe netfilter" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
