MPLS networks implementation using FRRouting

This repository includes three MPLS implementations using FRR routers. The first lab sets the MPLS labels manually in three-router topology. In the second lab, the labels are distributed using LDP (Label Distribution Protocol) in three-router ISP network connecting two customer routers. The third lab uses Virtual Routing and Forwarding (VRF) and BGP to create L3 VPN tunnels across an ISP network.

Requirements

To use this lab, you need to install containerlab (I used the script method Ubuntu 20.04 VM). You also need to have basic familiarity with Docker.

Also make sure MPLS is enabled on the host machine (where containerlab is installed). See below.

For troubleshooting and traffic inspection, you may also install Wireshark and/or Tshark in the host machine/VM.

Finally, clone this repository to your host VM:

git clone https://github.com/martimy/clab_mpls_frr

Enabling MPLS

To check if mpls is enabled on the host machine:

$ lsmod | grep mpls
mpls_gso               16384  0
mpls_iptunnel          20480  0
mpls_router            40960  1 mpls_iptunnel
ip_tunnel              24576  1 mpls_router

If you don't see the output above, load the modules and try again:

modprobe mpls_router
modprobe mpls_gso
modprobe mpls_iptunnel

To load the modules at boot time, add the following lines to /etc/modules-load.d/modules.conf:

cat >/etc/modules-load.d/modules.conf <<EOF
mpls_router
mpls_gso
mpls_iptunnel
EOF

Note also that MPLS needs to be enabled on each router. Therefore, lines similar to following are added to configuration of each router.

sysctl -w net.mpls.conf.lo.input=1
sysctl -w net.mpls.conf.eth1.input=1
sysctl -w net.mpls.platform_labels=1048575

Lab1 - Manual labels

This lab represents a network topology of three routers and two hosts attached to the end routers. To establish connectivity between the end hosts, MPLS labels are pushed popped or swapped manually in each router.

Use the following command to start the lab:

cd mpls_frr_static
sudo clab deploy --topo mpls-frr-static.clab.yml

To end the lab:

sudo clab destroy --topo mpls-frr-static.clab.yml

To access the CLI of an FRR router:

docker exec -it clab-mpls_frr_ldp-r1 vtysh

To access a host:

docker exec -it clab-mpls_frr_ldp-host4 bash

Try this

Test end-to-end connectivity between the two hosts:

docker exec -it clab-mpls_frr_static-host4 ping 192.168.2.5

Use tshark to inspect the MPLS labels at different interfaces (notice the use of network namespace):

sudo ip netns exec clab-mpls_frr_static-r2 tshark -i eth1 -O ethernet

The output should be similar to the following (notice the labels in each direction):

Frame 19: 102 bytes on wire (816 bits), 102 bytes captured (816 bits) on interface eth1, id 0
Ethernet II, Src: aa:c1:ab:58:60:3e (aa:c1:ab:58:60:3e), Dst: aa:c1:ab:3d:2d:9d (aa:c1:ab:3d:2d:9d)
MultiProtocol Label Switching Header, Label: 200, Exp: 0, S: 1, TTL: 62
Internet Protocol Version 4, Src: 192.168.1.4, Dst: 192.168.2.5
Internet Control Message Protocol

Frame 20: 102 bytes on wire (816 bits), 102 bytes captured (816 bits) on interface eth1, id 0
Ethernet II, Src: aa:c1:ab:3d:2d:9d (aa:c1:ab:3d:2d:9d), Dst: aa:c1:ab:58:60:3e (aa:c1:ab:58:60:3e)
MultiProtocol Label Switching Header, Label: 300, Exp: 0, S: 1, TTL: 63
Internet Protocol Version 4, Src: 192.168.2.5, Dst: 192.168.1.4
Internet Control Message Protocol

Lab2 - LDP

This lab represents a network topology of three ISP routers connecting two customer routers, each connected to a singe host. To establish connectivity between the end hosts, LDP is used to dynamically assign MPLS labels. OSPF is used among the ISP routers to advertise connected networks as well as static router to customer networks.

Use the following command to start the lab:

cd mpls_frr_ldp
sudo clab deploy --topo mpls-frr-ldp.clab.yml

To end the lab:

sudo clab destroy --topo mpls-frr-ldp.clab.yml

Try this

Test end-to-end connectivity between the two hosts:

docker exec -it clab-mpls_frr_static-host4 ping 192.168.2.5

Use tshark to inspect the traffic at different interfaces:

sudo ip netns exec clab-mpls_frr_ldp-r1 tshark -i eth1 -O ethernet

Selected output (echo/echo reply):

Frame 10: 102 bytes on wire (816 bits), 102 bytes captured (816 bits) on interface eth1, id 0
Ethernet II, Src: aa:c1:ab:64:e5:a0 (aa:c1:ab:64:e5:a0), Dst: aa:c1:ab:ac:67:1e (aa:c1:ab:ac:67:1e)
MultiProtocol Label Switching Header, Label: 16, Exp: 0, S: 1, TTL: 61
Internet Protocol Version 4, Src: 192.168.2.5, Dst: 192.168.1.4
Internet Control Message Protocol

Frame 11: 102 bytes on wire (816 bits), 102 bytes captured (816 bits) on interface eth1, id 0
Ethernet II, Src: aa:c1:ab:ac:67:1e (aa:c1:ab:ac:67:1e), Dst: aa:c1:ab:64:e5:a0 (aa:c1:ab:64:e5:a0)
MultiProtocol Label Switching Header, Label: 21, Exp: 0, S: 1, TTL: 62
Internet Protocol Version 4, Src: 192.168.1.4, Dst: 192.168.2.5
Internet Control Message Protocol

Selected output (OSPF and LDP exchanges):

Frame 15: 82 bytes on wire (656 bits), 82 bytes captured (656 bits) on interface eth1, id 0
Ethernet II, Src: aa:c1:ab:ac:67:1e (aa:c1:ab:ac:67:1e), Dst: IPv4mcast_05 (01:00:5e:00:00:05)
Internet Protocol Version 4, Src: 10.0.0.1, Dst: 224.0.0.5
Open Shortest Path First

Frame 16: 82 bytes on wire (656 bits), 82 bytes captured (656 bits) on interface eth1, id 0
Ethernet II, Src: aa:c1:ab:64:e5:a0 (aa:c1:ab:64:e5:a0), Dst: IPv4mcast_05 (01:00:5e:00:00:05)
Internet Protocol Version 4, Src: 10.0.0.2, Dst: 224.0.0.5
Open Shortest Path First

Frame 17: 84 bytes on wire (672 bits), 84 bytes captured (672 bits) on interface eth1, id 0
Ethernet II, Src: aa:c1:ab:ac:67:1e (aa:c1:ab:ac:67:1e), Dst: IPv4mcast_02 (01:00:5e:00:00:02)
Internet Protocol Version 4, Src: 10.0.0.1, Dst: 224.0.0.2
User Datagram Protocol, Src Port: 646, Dst Port: 646
Label Distribution Protocol

Lab3 - VRF

This lab demonstrates how an ISP can connect two customers with an overlapping IP address space using VRFs.

Running this lab is a bit tricky because of a bug in the FRR, so you must do in stages:

Initially, make sure the interface configuration in the starting configuration files (e.g. config/r1/frr.conf) are not assigned to a VFR.

interface eth2
  ...

Use the following command to start the lab:

cd mpls_frr_vrf
sudo clab deploy --topo mpls-frr-vrf.clab.yml

Test direct connectivity between all nodes:

./test-direct-connectivity.sh

Once ping is confirmed between each two adjacent routers, setup MPLS and VRF using the following script which enables MPLS, create VRFs and assign interfaces to them in the ISP routers.

./setup_mpls_vrf.sh

There is a bug in FRR that causes changes to route distinguishers and route targets commands in the BGP configuration, so make sure you change the BGP configuration in routers R1 and R3 as follows:

router bgp 100 vrf blue
  ...
  address-family ipv4 unicast
    ...
    rd vpn export 100:1
    rt vpn both 100:1
    ...
!
router bgp 100 vrf red
  ...
  address-family ipv4 unicast
    ...
    rd vpn export 100:2
    rt vpn both 100:2
    ...

After, the change, verify the routers in routers R1 and R3. They should be as follows:

$ docker exec -it clab-mpls_frr_vrf-r1 vtysh -c "show ip route vrf all"

Codes: K - kernel route, C - connected, S - static, R - RIP,
       O - OSPF, I - IS-IS, B - BGP, E - EIGRP, N - NHRP,
       T - Table, v - VNC, V - VNC-Direct, A - Babel, F - PBR,
       f - OpenFabric,
       > - selected route, * - FIB route, q - queued, r - rejected, b - backup
       t - trapped, o - offload failure

VRF blue:
C>* 10.0.2.0/24 is directly connected, eth2, 00:08:44
B>  10.0.3.0/24 [20/0] via 3.3.3.3 (vrf default) (recursive), label 80, weight 1, 00:04:30
  *                      via 10.0.0.2, eth1 (vrf default), label 17/80, weight 1, 00:04:30
S>* 192.168.1.0/24 [1/0] via 10.0.2.4, eth2, weight 1, 00:08:44
B>  192.168.2.0/24 [20/0] via 3.3.3.3 (vrf default) (recursive), label 80, weight 1, 00:04:30
  *                         via 10.0.0.2, eth1 (vrf default), label 17/80, weight 1, 00:04:30

VRF default:
K>* 0.0.0.0/0 [0/0] via 172.20.20.1, eth0, 00:08:45
O   1.1.1.1/32 [110/0] is directly connected, lo, weight 1, 00:08:44
C>* 1.1.1.1/32 is directly connected, lo, 00:08:44
O>* 2.2.2.2/32 [110/10] via 10.0.0.2, eth1, weight 1, 00:07:59
O>* 3.3.3.3/32 [110/20] via 10.0.0.2, eth1, label 17, weight 1, 00:07:49
O   10.0.0.0/24 [110/10] is directly connected, eth1, weight 1, 00:08:03
C>* 10.0.0.0/24 is directly connected, eth1, 00:08:44
O>* 10.0.1.0/24 [110/20] via 10.0.0.2, eth1, label implicit-null, weight 1, 00:07:59
O   172.20.20.0/24 [110/20] via 10.0.0.2, eth1, weight 1, 00:07:58
C>* 172.20.20.0/24 is directly connected, eth0, 00:08:45

VRF red:
C>* 10.0.4.0/24 is directly connected, eth3, 00:08:44
B>  10.0.5.0/24 [20/0] via 3.3.3.3 (vrf default) (recursive), label 81, weight 1, 00:04:30
  *                      via 10.0.0.2, eth1 (vrf default), label 17/81, weight 1, 00:04:30
S>* 192.168.1.0/24 [1/0] via 10.0.4.6, eth3, weight 1, 00:08:43
B>  192.168.2.0/24 [20/12] via 3.3.3.3 (vrf default) (recursive), label 81, weight 1, 00:04:30
  *                          via 10.0.0.2, eth1 (vrf default), label 17/81, weight 1, 00:04:30

Try this

Monitor the traffic at any link in the network using Wireshark. If Wireshark is installed in the local machine, use:

sudo ip netns exec clab-mpls_frr_static-r1 wireshark

To use Wireshark from a remote machine, use:

ssh -p 2222 user@remotehost "sudo ip netns exec clab-mpls_frr_static-r1 tcpdump -U -nni eth1 -w -" | wireshark -k -i -

The ssh command establishes a connection to is remotehost using port 2222 (optional, instead of the standard port 22) using user credentials, then run tcpdump on the remotehost (select the desired router and interface):

sudo ip netns exec clab-mpls_frr_static-r1 tcpdump -U -nni eth1 -w -

tcpdump options:

-U Print undecoded NFS (Network File System) handles.
-nn : do not resolve hostnames or ports.
-i eth1 : capture packets on interface eth1
-w - Write the raw packets to the standard output

The output of the tcpdump is piped to Wireshark (running locally), with these options:

-k start capturing immediately
-i interface name -

License

martimy/clab_mpls_frr

ErrorLooks like something went wrong!

About

Topics

Resources

License

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Languages

Error
Looks like something went wrong!

Packages