[c-nsp] ISP / MPLS "POP" design

Aaron aaron1 at gvtc.com
Wed Oct 30 00:27:37 EDT 2013

i contain my internet uplinks and my customer public subnet interfaces on
customer facing routers all within a vrf.

The router to router connections throughout my network core are all mpls
enabled and are 10.x.x.x addressed...and those all reside in what is known
as the default/global vrf (really no vrf at all, just the core igp table)

So again, all the public ip space and internet uplinks are all contained
within a vrf.

Throughout my network I have all my pe's (customer facing pe's and internet
facing dual pe's) all have 2, yes only 2, bgp neighbor sessions.  One to rr
cluster node 1 and the other to rr cluster node 2.  (those 2 rr cluster
nodes are of course what makes up my single rr cluster).


-----Original Message-----
From: cisco-nsp [mailto:cisco-nsp-bounces at puck.nether.net] On Behalf Of
CiscoNSP List
Sent: Tuesday, October 29, 2013 2:54 PM
To: cisco-nsp at puck.nether.net
Subject: [c-nsp] ISP / MPLS "POP" design


Hoping someone can provide some "best practice" advise on an ISP/MPLS

Customer's would have a mix of vrf and Internet tails

POP "A" has 2 x 7200's (PE's) and 2 x 6500's (Ps') - Both 7200's have a
single IPTransit service(Full BGP table) to different upstreams.

7200's and 6500's would be connected (physically) via a full mesh, with MPLS
enabled on p-t-p links

IGP(OSPF) on all p-t-p links, and only carry loopback IP's and p-t-p link
IP's iBGP under VPNv4 for MP-BGP (vrf IP's) with the 2 6500's as route

The part Im unsure of is the "global" iBGP (For customer "Internet" IP's) -
Would the 6500's also be route reflectors here? And with regards to the full
BGP tables on the 7200's, would the 7200's only send default route to the

Any suggestions/comments are greatly appreciated.

cisco-nsp mailing list  cisco-nsp at puck.nether.net
archive at http://puck.nether.net/pipermail/cisco-nsp/

More information about the cisco-nsp mailing list