[c-nsp] Number of route reflectors, best practice?

[Mack McBride]

The N squared problem dictates that centralized route reflectors (or confederations) 
become a necessity at some point.  Based on the information provided 6 full mesh devices
seems fairly good.  Even 9 full mesh devices is not too bad.  We currently run 14 cores 
in one of our backbones with a similar configuration (7600 cores as reflectors) and we do 
occasionally see CPU load issues (RSP720s).  However we are running multiple borders with full table
peers at a number of sites.  We are planning on adding another layer of route 
reflection (ASR1K) as we integrate our islands. Running much higher than 20 route reflectors 
(with wimpy processors) in a layer just seems to be asking for trouble.

Mack

-----Original Message-----
From: cisco-nsp-bounces at puck.nether.net [mailto:cisco-nsp-bounces at puck.nether.net] On Behalf Of Gert Doering
Sent: Thursday, July 21, 2011 2:22 PM
To: Peter Rathlev
Cc: cisco-nsp
Subject: Re: [c-nsp] Number of route reflectors, best practice?

Hi,

On Thu, Jul 21, 2011 at 08:56:13AM +0200, Peter Rathlev wrote:
> Is there another/better way of addressing this problem than adding 
> extra redundancy?
[..]
> We have a very small routing table at the moment (~10k prefixes, 
> largest VRF has ~2k, no Internet), and I'd gladly sacrifice some 
> convergence time for extra stability.

"Why have central RRs at all"?

"Just because everybody else does it" is a no-go in my book :-) - we currently have a design similar to your current design, that is, all "core" routers (8) are full-meshed, and all "edge" routers in a given POP use the "core" as RRs.  Edges have only edge-routes plus default, so the computational effort on the RRs is not that bad.  And we don't need extra boxes...

If both cores fail in a POP, that POP is down anyway, and I don't need to worry about RR reachability either.

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