[c-nsp] Rationale for ISIS default origination behavior

[Saku Ytti]

On (2013-01-22 17:50 +0100), Gert Doering wrote:

> I'm still not convinced that this is more elegant than "the to-upstream
> edge routers just inject a static default route (pointing to the upstream
> peer) into IGP".  What's the benefit of the extra recursion?

If you point it at peer IP, it'll be valid route, as long as peer interface
is UP. This does not mean it's usable egress.

If at edge you point it to some 8.8.8.8 without interface and send
next-hop-self, it'll advertise it, even if local edge is down as it'll
recurse to any available edge presuming edge has full table.

If you do route to 8.8.8.8 with explicit WAN interface it's mostly the
same, but I think some platforms I've tried stop recursing when interface
is given.
And if it works in your platform and you can choose option with same
results, where one gives you default free BGP and one does not, for me,
choice is easy.

-- 
  ++ytti