[c-nsp] Internet inside a VRF?

[Vitkovsky, Adam]

I guess you can ask: Why do we run mpls anyway or even plan on expanding it all the way to the access layer right? 
I thought the answer is obvious, TE capabilities, fast failover or common carrier infrastructure that scales well 
And by common I mean infrastructure that supports all the services you offer not just a couple

In pure ipv4 environment how do you make sure your RRs infrastructure offers multiple paths for particular prefix? -well with mpls we just configure each PE with a different RD for the internet VRF -I know of several solutions to this in pure ipv4 but none is this simple

adam

-----Original Message-----
From: cisco-nsp-bounces at puck.nether.net [mailto:cisco-nsp-bounces at puck.nether.net] On Behalf Of Gert Doering
Sent: Wednesday, March 14, 2012 10:04 AM
To: Pshem Kowalczyk
Cc: cisco-nsp at puck.nether.net
Subject: Re: [c-nsp] Internet inside a VRF?

Hi,

On Wed, Mar 14, 2012 at 01:12:02PM +1300, Pshem Kowalczyk wrote:
> In my previous role we've done just that. One internet VRF for all
> transit functions, separate vrfs for peering and customers and
> import-export statements to tie them all together. 

What is the benefit?  The obvious drawback is "much more complicated,
more possible ways things can blow up, and more effort to setup and
maintain".

(We currently run Internet in the global table, and do not currently
intent to change that due to "if we make it more complicated, people
will make more interesting mistakes")

gert
-- 
USENET is *not* the non-clickable part of WWW!
                                                           //www.muc.de/~gert/
Gert Doering - Munich, Germany                             gert at greenie.muc.de
fax: +49-89-35655025                        gert at net.informatik.tu-muenchen.de