Configure MPLS VPNv4 Routes on Cisco IOS Step-by-Step

Hello! This is Lazarus at telecomTech, where  telecom and networking technologies... are simply   explained! Today we'll be simply explaining  the configuration of VPNv4 routes on the MPLS   backbone network using Cisco IOS routers. Now  MPLS can become quite complex and for this reason,  

this whole video series starts from scratch. If  you haven't already done so, check out the first   videos in this series allowing you to gain the  required understanding of the fundamentals in   order to more readily appreciate the information  shared in this video. And remember to subscribe,  

enable notifications, and like this video if you  like this video. And if you have questions or   feedback, please write them in the comments below!  I'm always happy to read them and to respond! The   topology that we'll be looking at today is  this one here. For this video, we're going  

to focus only on the MPLS backbone infrastructure  itself. We're going to leave the PE-CE interaction   and relationship for the next video. Let  me explain a few things about what we're   going to do. Although in a production network you  would obviously configure all of these routers,  

for our purposes we'll only be configuring these  four routers here, which are the routers that are   going to be involved in the exchange of routes  and the label switching for the networks of these   two customer sites. First we'll be looking at  configuring the underlay network of the PE and  

P routers by creating loopback interfaces and  assigning them IP addresses for the purpose of   forming LDP adjacencies. We'll also assign an  IP addressing scheme to the physical interfaces   as well. Next we'll employ OSPF between all  PE and P routers including the loopbacks,  

except for those interfaces facing the customers,  such as these ones here. Third we'll enable MPLS   on all the appropriate interfaces on the P and  PE routers. Fourth, we'll create the appropriate   VRFs, RDs, and RTs in the PE routers. And finally  we'll create the iBGP peerings between PE1 and  

PE2 using multi-protocol BGP so that the VPNv4  routes will be advertised between them. Let's   get started! Here is our underlay network topology  and the information with our desired IP addressing   scheme. Let's see what our configuration steps  are for the PE1 router here based on these IP  

addresses and interfaces. We start by creating the  loopback0 interface and assign it the 1.1.1.1/32   address. We then assign the appropriate IP address  to the GigabitEthernet0/2 interface which is the   interface that faces into the MPLS network and  connects to this P router here. We then configure  

our GigabitEthernet0/1 interface which is the  interface that faces the customer CE router. And   finally we configure OSPF such that the loopback  address as well as the network connected to the   GigabitEthernet0/2 interface are being advertised.  We don't advertise the network connected to the  

CE router, however. Now here we've configured  this GigabitEthernet0/1 interface as passive.   This isn't actually obligatory, but it is best  practice, so that OSPF will not send out hello   messages to the CE router in an attempt to create  an adjacency. Next, let's see what configurations  

we're going to have on this P router here. As you  can see, they're similar to the PE1 router. First   we configure the loopback, then the IP addresses  of the GigabitEthernet interfaces. Then we   configure OSPF and make sure that we advertise the  networks of all interfaces including the loopback  

interface. Remember, OSPF must advertise the  loopbacks between the PE and P routers in order   for them to establish LDP adjacencies. Now the  configurations for the underlay of the P2 and   PE2 routers are very similar, but we'll show  them quickly here for completeness. You can see  

the GigabitEthernet interfaces are configured  with IP addresses as is the loopback interface   and all of those are advertised using OSPF.  And the PE2 router here has its loopback and   physical interfaces configured with IP addresses  and has its core-network-facing interface and its  

loopback interface advertised by OSPF, but not its  customer-facing interface. And once again, as best   practice, we have set this interface as passive.  So now we've set up our IP addressing scheme,   our loopbacks, and our OSPF IGP underlay routing.  Now the next step is to enable MPLS and the LDP  

protocol. To do that we simply have to enable MPLS  on all those interfaces through which the P and PE   routers will create LDP adjacencies. That means  all non-customer-connecting interfaces on the   PEs and all physical interfaces of the P routers.  Here you can see that we have enabled MPLS using  

the `mpls ip` command on the Gigabit 0/2 interface  of PE1 and also on the GigabitEthernet0/2 and 0/3   interfaces of the P1 router. The same is done on  the P2 and the PE2 routers like so. Νow we have   set up the underlay network, the IP addressing,  and MPLS, allowing the P and PE routers to create  

LDP adjacencies. Now we go on to create our  VRFs in our PEs. Remember, VRFs are used to   identify specific customers and are applied only  to PEs. Let's see what we will configure in our   PE1 router. First, we create our VRF, and let's  call it GREEN. When you issue the `ip vrf green`  

command, you can see that we enter into the VRF  configuration mode, and it's indicated by the   `config-vrf` indicator here. Now it's here that  we configure the route distinguisher for this VRF,   and it is the RD that will create our unique  VPNv4 route. Using our context sensitive help,  

we issue the `rd ?` command, and we can see the  options that are available for us. We can add   our VPN route distinguisher using either the  format of ASN:NN or IP address:NN. Remember,   the RD is a 64-bit value, so the number before  the colon is a 32-bit number and the number  

after the colon is another 32-bit number. So you  are given the option of using the dotted decimal   format of an IP address, which is a 32-bit value,  or the value of the ASN that you're using in your   BGP configuration. Now note that these are just  suggested formats to make our lives easier. You  

can use any values you like, but it may be useful  to associate the RD value somehow with an ASN, or   an IP address for easier management. In any case,  I will be using the 100:1 value since we're going   to be using ASN 100 in our MPLS infrastructure  when we configure BGP. Now at the same time, under  

the VRF configuration mode, we also configure  our route target, and we configure it like so.   Now this doesn't have to be the same value as the  RD, but I'll keep it the same just for uniformity.   Remember the RT is used to import and export VPNv4  routes to and from MP-BGP. You can use different  

RTs for importing and exporting but I will use  the same for both. So now we have configured our   VRF and our associated RTs and RDs. Next we have  to associate the customer's interface with this   VRF. We do so by going into the configuration  mode of our Gigabit 0/1 interface which is the  

customer-facing interface on our PE, and we assign  our VRF there like so. Now when we do this, we get   a Syslog message telling us that the IP address  we had originally configured on the interface   has been removed. This is normal behavior when  you assign a VRF to an interface that already  

has an IP address. So we have to configure it  again. And that's it for PE1! Let's quickly see   the corresponding configuration on PE2. Now the  RDs can be, but don't have to be, the same on the   PEs since they are used to uniquely identify the  networks on the locally connected CE. Conversely,  

the RTs should be the same on both ends so that  they are exported to the appropriate customer   VRF when advertised. To keep things tidy, it  is always best practice to keep them the same   so you can easily identify to which customer they  belong. Now finally, we come to the configuration  

of our multi-protocol BGP, where we will configure  an iBGP peering between the PE routers so they can   share the VPNv4 routes. We start off with the iBGP  peerings. Let's take a look at the configuration   in PE1. We do so with our normal BGP configuration  using the local AS of 100 and the iBGP neighbor of  

4.4.4.4 which is PE2. We also set the source of  our iBGP updates to the local loopback interface.   Now this is where multi-protocol BGP comes in.  Here we configure our BGP peering to share VPNv4   routes using the address family VPNv4, and we  activate PE2 as our neighbor for sharing these  

routes. Now this will automatically enable another  feature of BGP. Take a look at the resulting   configuration of BGP in PE1. We see that the PE  router automatically added the `send-community   extended` feature. This is necessary because our  RTs are shared using extended BGP communities.  

Now to complete our configuration, let's take  a look at the counterpart configuration on our   PE2 router. Here we see the iBGP peering with  PE1 is configured and we also enable the VPNv4   address family so that iBGP is able to share  those VPNv4 routes which are composed of the  

IP address of the destination prepended with that  RD value, making the VPNv4 routes unique on the   MPLS infrastructure. So at this point, we have  completed our configuration of the ISP network for   VPNv4 routes using MPLS on Cisco IOS routers. In  the next video, we'll continue this configuration  

with the setup of the customer devices. We'll  also go through several verification steps to see   how this whole setup operates in action, so stay  tuned for the next video! I hope you found this   video useful and if you have please click that  thumbs-up button. If you have any questions feel  

free to leave a comment below and subscribe to get  updates to newly published videos. I'm Lazarus at   telecomTech, I hope this has been helpful for  you and I'd like to thank you for watching!