[c-nsp] small (tiny?) route reflector for MPLS L3VPN network

[Phil Mayers]

We run a network of 6500s and Juniper J4350s with OSPF IGP and iBGP 
including vpnv4 (with mdt address family) and vpnv6.

Two of the "central" 6500s have served as the route reflectors for a 
while now without problem, however there is a question over whether we 
could usefully deploy a pair of dedicated boxes. For one thing it would 
mean we could reboot traffic-forwarding devices without affecting BGP 
sessions, for another we could get onto a "newer" IOS with features like 
BGP "selective address tracking" to kill BGP sessions on IGP flaps.

We have a pretty small routing table:

#sh ip route summary
IP routing table name is Default-IP-Routing-Table(0)
IP routing table maximum-paths is 32
Route Source    Networks    Subnets     Overhead    Memory (bytes)
connected       1           20          1580        3024
static          16          14          2232        12496
ospf 1          2           66          5328        9792
   Intra-area: 54 Inter-area: 0 External-1: 0 External-2: 14
   NSSA External-1: 0 NSSA External-2: 0
bgp 64580       11          923         67320       134496
   External: 17 Internal: 917 Local: 0
internal        25                                  54700
Total           55          1023        76460       214508
Removing Queue Size 0

...although each VRF contains the whole table, so it's actually 10x that.

I think a 2900 (or even a 1900) with advanced IP services will more than 
suffice for this. Any comments?