[130353] in North American Network Operators' Group

home help back first fref pref prev next nref lref last post

Re: RIP Justification

daemon@ATHENA.MIT.EDU (Tim Franklin)
Fri Oct 1 12:28:39 2010

Date: Fri, 1 Oct 2010 16:28:30 +0000 (GMT)
From: Tim Franklin <tim@pelican.org>
To: nanog@nanog.org
In-Reply-To: <28460806.51285950096710.JavaMail.root@jennyfur.pelican.org>
Errors-To: nanog-bounces+nanog.discuss=bloom-picayune.mit.edu@nanog.org

----- "Ruben Guerra" <Ruben.Guerra@arrisi.com> wrote:

> Using BGP would be overkill for most. Many small commercial customers
> to not want the complexity of BGP

This one keeps coming up.

Leaf-node BGP config is utterly trivial, and is much easier for the SP to c=
onfigure the necessary safety devices on their side to stop you from shooti=
ng yourself in the foot and blowing up your networks - or worse, *their* ne=
twork.  Plus, if / when in the future you need to do something clever, you'=
ve already got the routing protocol with all the advanced knobs in place, r=
eady for you to tweak as needed.

The Enterprise guys really need to get out of the blanket "BGP is scary" mi=
ndset - running BGP for an SP with multitudes of customers, peers, transits=
, aggregation, filters etc and getting it right needs expertise and experie=
nce.  Announcing a /24 LAN and receiving a default on a single link, not so=
 much.

> or want to spend money on extra
> resources (routers that actually support it)

This has a bit more weight to it, if you're at the really low end (certainl=
y the consumer end).  But a BGP-capable Cisco 800-series is, what, =C2=A330=
0?

Regards,
Tim.


home help back first fref pref prev next nref lref last post