vrf... the road to mpls...

this is the beginning of what would later bury frame relay...

notice the diagram... the ip addresses are the same for the clients... virtual route forwarding... see below

r1#ping 10.1.1.1                                                                                     
                                                                                                     
Type escape sequence to abort.                                                                       
Sending 5, 100-byte ICMP Echos to 10.1.1.1, timeout is 2 seconds:                                    
!!!!!                                                                                                
Success rate is 100 percent (5/5), round-trip min/avg/max = 1/2/4 ms                                 
r1#

r2#ping 10.1.1.1

Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 10.1.1.1, timeout is 2 seconds:
!!!!!
Success rate is 100 percent (5/5), round-trip min/avg/max = 4/4/4 ms
r2#

r4#ping 10.1.1.1

Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 10.1.1.1, timeout is 2 seconds:
!!!!!
Success rate is 100 percent (5/5), round-trip min/avg/max = 4/4/4 ms
r4#

each client router (ce, client edge) is connected to  r3 (pe, provider edge) with a virtual connection provided by vrf... like frame relay without the hassle of a frame switch... this would later give way to mpls, or frame tagging for the wan...

normally, if you try to assign the same addresses to interfaces the router complains you have overlapping ip's... with a little variable magic, and very little at that, you can change the game...

r3#ping vrf r1 10.1.1.1

Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 10.1.1.1, timeout is 2 seconds:
!!!!!
Success rate is 100 percent (5/5), round-trip min/avg/max = 1/2/4 ms
r3#ping vrf r2 10.1.1.1

Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 10.1.1.1, timeout is 2 seconds:
!!!!!
Success rate is 100 percent (5/5), round-trip min/avg/max = 4/4/4 ms
r3#ping vrf r4 10.1.1.1

Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 10.1.1.1, timeout is 2 seconds:
!!!!!
Success rate is 100 percent (5/5), round-trip min/avg/max = 4/5/8 ms
r3#

first note the rd's (route-distinguishers) in the diagram in global config...

then assign them to the interfaces...

interface FastEthernet0/0
 ip vrf forwarding r1
 ip address 10.1.1.1 255.255.255.0
 speed 100
 full-duplex
!
interface Serial0/0
 ip vrf forwarding r2
 ip address 10.1.1.1 255.255.255.0
 encapsulation ppp
 no fair-queue
 clock rate 1000000
!
interface Serial0/1
 ip vrf forwarding r4
 ip address 10.1.1.1 255.255.255.0
 encapsulation ppp
 clock rate 1000000

of course, it's just the beginning... you are running a separate route table for each vrf... see below

look ma, no global route table...

r3#sh ip route
Codes: C - connected, S - static, R - RIP, M - mobile, B - BGP
       D - EIGRP, EX - EIGRP external, O - OSPF, IA - OSPF inter area
       N1 - OSPF NSSA external type 1, N2 - OSPF NSSA external type 2
       E1 - OSPF external type 1, E2 - OSPF external type 2
       i - IS-IS, su - IS-IS summary, L1 - IS-IS level-1, L2 - IS-IS level-2
       ia - IS-IS inter area, * - candidate default, U - per-user static route
       o - ODR, P - periodic downloaded static route

Gateway of last resort is not set

r3#

then...

r3#sh ip route vrf r4

Routing Table: r4
Codes: C - connected, S - static, R - RIP, M - mobile, B - BGP
       D - EIGRP, EX - EIGRP external, O - OSPF, IA - OSPF inter area
       N1 - OSPF NSSA external type 1, N2 - OSPF NSSA external type 2
       E1 - OSPF external type 1, E2 - OSPF external type 2
       i - IS-IS, su - IS-IS summary, L1 - IS-IS level-1, L2 - IS-IS level-2
       ia - IS-IS inter area, * - candidate default, U - per-user static route
       o - ODR, P - periodic downloaded static route

Gateway of last resort is not set

     10.0.0.0/8 is variably subnetted, 2 subnets, 2 masks
C       10.1.1.2/32 is directly connected, Serial0/1
C       10.1.1.0/24 is directly connected, Serial0/1
r3#

so what's the big deal with mpls... it doesn't have to dig into the packet, it labels the packet, it tags it, and uses cef to switch the hell out of it...