IPv6 Internet In VRF Over MPLS

In the previous post, I have used 6PE concept to access the IPv6 Internet over IPv4 MPLS VPN cloud through global routing table. But this post is little bit different from the previous one. In this post, I will be using IPv6 Internet VRF to serve the IPv6 internet over IPv4 MPLS cloud. The main advantage of using this model is that the entire IPv6 internet route will be a part of vrf but in IPv6 from MPLS post entire ipv6 traffic was a part of global route table. Provisioning of serving IPv6 internet will remain same as it happens in IPv4. For IPv4 internet provisioning in vrf please refer to post Provisioning Internet with Internet VRF. The only difference is that in IPv6 we need to use the address family VPNv6 and IPv6. Given depicted images shows the graphical flow of serving IPv6 internet in VRF over IPv4 MPLS cloud.

Figure 1 is depicted about the default route announcement in IPv6 Internet VRF and the same route is being forwarded by label 100 to the remote PE.

Figure 1

Figure 2 is depicting about the CPE IPv6 prefixes in IPv6 Internet VRF with label 200.

Figure 2

Figure 3 is depicting about the forwarding traffic in IPv6 Internet VRF from CPE to PE.

Figure 3

Figure 4 is depicting about the reverse traffic from IPv6 Internet cloud to IPv6 Internet VRF.

Figure 4

26DC33AXBVX7


Click Here To Read Rest Of The Post...

IPv6 Addressing Architecture For Service Providers

ARIN has given IPv6 addressing plan for service providers. The points of IPv6 allocation is given below:-

1. Separate address block for infrastructure from other uses (enterprise, loopbacks),it may be two /48s per PoP.
2. Each individual site should receive a /48 assignment which 65536 subnets (2^16).
3. Any prefixes shorter than /48 will only be assigned when there is written justification to show that this prefix will meet the RIR HD ratio guidelines within 5 years.
4. Each PoP is a site therefore assign a /48 for infrastructure.
5. No subnets will use prefixes longer than /64.
6. Assign a /64 per LAN / VLAN / subnet.
7. The enterprise network should receive a prefix sufficient to provide a /48 allocation for each site (office/campus/PoP) at which the company has employees or systems.
8. If you need private addresses, generate a ULA prefix as defined in RFC 4193.
Source:- ARIN WIki

Click Here To Read Rest Of The Post...

IPv6 From MPLS

If the service provider is having MPLSVPN backbone and would like to serve IPv6 internet; In that case with the help of MP-iBGP this objective can be achieved with no infrastructure changes. This technique is also known as 6PE. Given depicted figure 1 shows the MPLS VPN ip4 cloud with ldp as label distribution protocol and OSPF is implemented as IGP. Edge routers PE1 and PE2 are configured as dual stack router which supports both IPv4 as well as IPv6 addressing. Both edge routers are configured with IPv6 - MP-iBGP with route reflector router. A default IPv6 route is configured on PE2 towards IPv6 Internet Cloud. PE2 is advertising default IPv6 route to RR with label 100 as next hop PE2 ipv4 BGP router id. The same default IPv6 route is installed on PE1 with outgoing IPv6 label 200 with PE2 BGP IPv4 router-id as next hop address.

Figure 1

In the given depicted figure 2, 6CE is wants to access the IPv6 Internet backbone. So a static point to point peering is configured with PE1. 6CE is having a default route towards PE1 and PE1 is having IPv6 route of 6CE loopback which is 2001::ABCD::1/64. The 6CE static route is advertised by PE1 to RR with label 200 as PE1 BGP IPv4 address as next hop address. The same route is being received by PE2.

Figure 2

In the given depicted figure 3,

Figure 3

It shows the pictorial view of forwarding traffic from 6CE to IPv6 Internet cloud. The traffic is first forwarded towards PE 1 which appends the IPv6 label 100 for destination 2001::1::1 and label 20 for the next hop address which is PE2 BGP router id. Then the traffic next moves to IPv4-P1 router which only swaps the topmost 20 label to 21 and forwards towards IPv4-P4/RR router. RR swaps the topmost label 21 to 22 and forwards to the next hop which is IPv4-P2. IPv4 performs the PHP (Penultimate Hop Popping) function and forwards the traffic with IPv6 label towards PE2. PE2 removes the all label and forwards the traffic to IPv6 internet cloud. The main advantage of using MP-iBGP is that the 6CE customers can access the IPv6 internet cloud by using the existing IPv4 cloud. By using LDP IPv6 route becomes hide as in legacy VPNv4, so the core routers doesn’t know about the IPv6 route and only look for the next hop address which is reachable by IGP.

Figure 4 shows the reverse traffic from IPv6 Internet cloud to 6CE.

Figure 4

By using the above mentioned technique, service provider need to use OSPF or IS-IS for IPv6. Also refer to my previous post about the selection of IGP protocol.
Click Here To Read Rest Of The Post...