MPLS introduction & Basic configuration
Multi protocol labels switching (MPLS) uses labels to forward packets, instead of traditional routing. Alternatively to having each router looking up the destination in each packet’s header and forwarding it to the next hop, routers can simply switch packets based label value.
Once a label distribution protocol (LDP) session is established between routers, labels are exchanged for prefixes and stored in the forwarding equivalence class table (FEC). A subset of packets to be forwarded along the same path and treatment.
Four bytes label is inserted between layer2 header and layer3 packet.
Router role:
- LSR Label Switch router a router which supports MPLS
- Ingress LSR Receives unlabeled packets and push a label on it.
- Egress LSR Pop a label from a packet.
- Intermediate LSR Switch labelled packets.
It is sometimes said that one of the benefits of MPLS is speeding up packet forwarding, since this operation no longer relies on routing table lookups. Although this might have been true in the old days, it no longer holds. Current routers architecture decouple data and control planes, offloading the CPU from the forwarding function.
So what are the real benefits and advantages of deploying MPLS ?
For starters, deploying MPLS will remove the need for running BGP on core nodes. This is referred to as BGP free core, where BGP is only running on edge nodes and often a set of route reflectors.
Another advantage is the possibility to implement VPN, over one AS or more (Inter-AS VPN). This design allows a given provider to segregate its routing context per VPN set. Traffic engineering (TE) is another benefit, whereby traffic path can be manipulated and engineered due to capacity constraints or links quality.
And finally AToM, which is an acronym for Any Transport over MPLS. Layer 2 VPNs (L2VPN) are a great application of AToM.
MPLS configuration
At a basic level, configuring MPLS is rather straight forward.
You must enable or have an IGP running already and enable LDP globally. Define or hard code LDP update source, this is typically default to a used Loopback zero. And finally, enable LDP on each interface adjacent to an LSR.
There is also the option to auto configure LDP on each interface running IGP.
!!! IOS based config !!!
interface GigabitEthernet0/1
ip address 10.1.1.1 255.255.255.254
ip router isis NL
mpls ip
!
router isis NL
net 01.0172.0016.0001.0001.00
is-type level-2-only
passive-interface Loopback0
!
mpls ldp router-id Loopback0!!! XR based config !!!
router isis NL
is-type level-2-only
net 01.0172.0016.0001.0004.00
address-family ipv4 unicast
mpls ldp auto-config
!
!
mpls ldp
router-id 172.16.1.4
!Notice how for XR config, I chose to auto-configures MPLS on any interfaces running that particular IGP process. Another difference between IOS and XR is the way router-id is configured.
For validation and verification following commands can be used.
!!! IOS based config !!!
sh mpls interfaces
Interface IP Tunnel BGP Static Operational
GigabitEthernet0/1 Yes (ldp) No No No Yes
GigabitEthernet0/2 Yes (ldp) No No No Yes
!
!
P2#sh mpls ldp neighbor
Peer LDP Ident: 172.16.1.2:0; Local LDP Ident 172.16.1.3:0
TCP connection: 172.16.1.2.646 - 172.16.1.3.54388
State: Oper; Msgs sent/rcvd: 1543/1542; Downstream
Up time: 22:14:24
LDP discovery sources:
GigabitEthernet0/1, Src IP addr: 10.1.1.3
Addresses bound to peer LDP Ident:
192.168.1.182 10.1.1.0 10.1.1.3 172.16.1.2
Peer LDP Ident: 172.16.1.4:0; Local LDP Ident 172.16.1.3:0
TCP connection: 172.16.1.4.58119 - 172.16.1.3.646
State: Oper; Msgs sent/rcvd: 34/33; Downstream
Up time: 00:18:08
LDP discovery sources:
GigabitEthernet0/2, Src IP addr: 10.1.1.5
Addresses bound to peer LDP Ident:
192.168.1.184 10.1.1.5 10.1.1.6 172.16.1.4
!
!
P2#sh mpls forwarding-table
Local Outgoing Prefix Bytes Label Outgoing Next Hop
Label Label or Tunnel Id Switched interface
16 Pop Label 172.16.1.4/32 0 Gi0/2 10.1.1.5
17 Pop Label 172.16.1.2/32 0 Gi0/1 10.1.1.3
19 Pop Label 10.1.1.0/31 1140 Gi0/1 10.1.1.3
20 16 172.16.1.1/32 0 Gi0/1 10.1.1.3
21 24005 172.16.1.5/32 0 Gi0/2 10.1.1.5
22 Pop Label 10.1.1.6/31 1140 Gi0/2 10.1.1.5
!
!
!!! XR based config !!!
RP/0/0/CPU0:P3#sh mpls ldp interface
Tue Jan 21 12:05:17.511 UTC
Interface GigabitEthernet0/0/0/0 (0x40)
VRF: 'default' (0x60000000)
Enabled via config: IGP Auto-config
Interface GigabitEthernet0/0/0/1 (0x60)
VRF: 'default' (0x60000000)
Enabled via config: IGP Auto-config
!
!
RP/0/0/CPU0:P3#sh mpls ldp neighbor
Tue Jan 21 12:06:05.398 UTC
Peer LDP Identifier: 172.16.1.3:0
TCP connection: 172.16.1.3:646 - 172.16.1.4:58119
Graceful Restart: No
Session Holdtime: 180 sec
State: Oper; Msgs sent/rcvd: 38/39; Downstream-Unsolicited
Up time: 00:22:22
LDP Discovery Sources:
IPv4: (1)
GigabitEthernet0/0/0/0
IPv6: (0)
Addresses bound to this peer:
IPv4: (4)
10.1.1.2 10.1.1.4 172.16.1.3 192.168.1.183
IPv6: (0)
Peer LDP Identifier: 172.16.1.5:0
TCP connection: 172.16.1.5:19855 - 172.16.1.4:646
Graceful Restart: No
Session Holdtime: 180 sec
State: Oper; Msgs sent/rcvd: 20/20; Downstream-Unsolicited
Up time: 00:06:42
LDP Discovery Sources:
IPv4: (1)
GigabitEthernet0/0/0/1
IPv6: (0)
Addresses bound to this peer:
IPv4: (3)
10.1.1.7 172.16.1.5 192.168.1.185
IPv6: (0)
!
!
RP/0/0/CPU0:P3#sh mpls forwarding
Tue Jan 21 12:06:16.937 UTC
Local Outgoing Prefix Outgoing Next Hop Bytes
Label Label or ID Interface Switched
------ ----------- ------------------ ------------ --------------- ------------
24000 Pop 172.16.1.3/32 Gi0/0/0/0 10.1.1.4 2450
24001 17 172.16.1.2/32 Gi0/0/0/0 10.1.1.4 0
24002 Pop 10.1.1.2/31 Gi0/0/0/0 10.1.1.4 0
24003 19 10.1.1.0/31 Gi0/0/0/0 10.1.1.4 1040
24004 20 172.16.1.1/32 Gi0/0/0/0 10.1.1.4 0
24005 Pop 172.16.1.5/32 Gi0/0/0/1 10.1.1.7 34848 In this post MPLS packet flow, you can learn more about label switching in an MPLS network. And follow the trail hop by hop, using Cisco show commands.
