> The "one interface per key" rule adds complexity but I think it also
> brings benefits by requiring such.

It's conceptually elegant and easy to explain.  These are good features to have.

> About the only issue I see for an interchange like this is that you'd
> also need one ipv4 address per veth, which are kind of scarce.

> unless there was a way to do the obscure extended nexthop idea in
> https://tools.ietf.org/rfc/rfc5549.txt in babel and linux?

Babel (at least babeld, don't know about BIRD) should deal fine with
Linux-style unnumbered links (which are different from Cisco-style
unnumbered links).

In the Linux style, you put the same IPv4 address on multiple links:

  ip addr add 192.0.2.1/32 dev eth0
  ip addr add 192.0.2.1/32 dev eth1
  ip addr add 192.0.2.1/32 dev wlan0

Now you run babeld on all interfaces, and everything should just work --
look, Ma, just one IPv4 address per host.

(In Cisco style, you put your IPv4 address on the loopback interface, and
the other interfaces magically borrow the loopback's IPv4 address.  I find
the Linux style easier to comprehend, but perhaps that's just me.)

-- Juliusz

_______________________________________________
Babel-users mailing list
[email protected]
https://alioth-lists.debian.net/cgi-bin/mailman/listinfo/babel-users

Reply via email to