[MLB-WIRELESS] Revisiting OSPF

Dan Flett conhoolio at hotmail.com
Thu Dec 16 11:23:08 EST 2004


Hi David,

> I don't think the issue of administrative independance is all that
great,
> in the context of being an ISP  on the internet you need isolation
from
> other AS routing policies whereas in MW there is a sort of agreed
policy
> (otherwise your not playing).  Probably a bigger concern here is
working
> out who needs to run a routing daemon and who doesn't.

Yes, I'm a fan of making it as simple as possible for people to run a
routing node.  I'm thinking that if a bunch of friends who live near to
each other all get connected, they can run their own interior routing
protocol amongst themselves and speak BGP to the rest of the network.
They don't have to if they don't want to - they could indeed each have
their own private ASN (Autonomous System Number for those at home).  The
advantage of running BGP instead of OSPF between areas is isolation -
OSPF can be destabilised fairly easily if someone misconfigures their
router.  If this happens in a regional AS the instability is limited to
that AS.  And the idea behind a multi-node regional AS is that one or
more persons would have root access to all the routers in that AS,
allowing problems to be quickly fixed.  If the people in the AS don't
trust each other to give each other root access, then they should split
into separate AS's (or not form a multi-node AS in the first place) and
speak BGP to each other.

> A node that has multiple links would need to run the agreed routing
> depending on how tightly the cooperation with the other end(s) ( It
may be
> BGP to both links or BGP to one and RIP to the other).

Running RIP suggests that both ends are using the same BGP ASN and are
deciding their own internal routing policy.  That's local independence -
whilst still cooperating with the overall Melbourne Wireless routing
policy.

> As suggested to you in the quagga forums, perhaps the best way to go
would
> be to treat each node as an AS. When nodes establish links they agree
to
> run BGP and communicate as neighbours, or they use OSPF / RIP / static
> routes and become a larger AS.

Exactly, and note that these Regional AS's may not necessarily be
directly aligned with the Melbourne Wireless Regional Groups as they
currently stand.  Multi-node AS's may form around clusters of local
connectivity - RGSouthern as it currently stands has a couple of such
clusters in it.
 
> This is where it gets messy. iBGP would mean each backbone node having
> each other backbone nodedefined as a neighbour (they would have the
same
> AS), there would potentially be hundreds of  nodes to the backbone.

At the recent BBQ the committee outlined a possible strategy for the
Melbourne Wireless Backbone - a true Backbone would be built and
administered directly by the MW committee - this would consist of about
four nodes in high-visibility locations - all directly linked to each
other, and with links into the regional clusters.

So in the medium-term, these four (or whatever number) "supernodes"
could be an AS unto themselves, running OSPF and iBGP internally, and
speaking eBGP to all other nodes.  If every other connected node in
Melbourne spoke only BGP to each other, it would still probably work.  I
now need to read up a bit more on BGP. :)

Cheers,

Dan

To unsubscribe: send mail to majordomo at wireless.org.au
with "unsubscribe melbwireless" in the body of the message



More information about the Melbwireless mailing list