[c-nsp] OSPF on Ring Networks

[Christian Meutes]

Hi,

On Sat, 14 May 2011 12:30:16 +0100, Nick Hilliard <nick at foobar.org> wrote:

> There is no "correct" answer to this question.  It depends completely on

> your traffic profile.  Probably the default ospf settings will do a 
> reasonable job of handling this situation.  And given that it's only 11 
> devices, there's no reason not to have them all in the same ospf area.
> 
> Other than that, you need to keep as few prefixes as possible in your
IGP 
> (preferably only loopbacks, with everything else carried in ibgp).  You 

Here it depends. If the next-hop changes OSPF will in most cases converge
faster than BGP. So if there aren't that many routes and you can live with
less policy possibilities I would prefer OSPF over BGP for carrying
non-infrastructure networks.

> should also probably drop the default hello/dead time from 10/40 seconds
> to 
> something a lot lower (e.g. 1/4 seconds), unless some of your routers
are 
> slow or busy.

BFD, MPLS-FRR, SPF- and Flood-Tuning for even faster convergence.

Cheers
Christian