WAN Technologies — MPLS and Metro-E

Multiprotocol Label Switching (MPLS) is a data plane technology that directs data from one network node to the next based on short path labels rather than long network addresses. It operates between Layer 2 and Layer 3, often called Layer 2.5. MPLS was developed to improve speed and scalability in core networks while supporting multiple protocols (IPv4, IPv6, etc.) and traffic engineering.

Traditional IP routing requires every router to perform a longest-prefix match lookup in the routing table for each packet, which can be slow and complex. MPLS adds a label to each packet, allowing routers to make forwarding decisions by simply swapping labels—a faster, simpler operation. Additionally, MPLS enables services like MPLS VPNs (L3VPN and L2VPN) and Traffic Engineering (TE).

- LSR (Label Switch Router): A router that performs MPLS forwarding. There are three types: - Ingress LSR: The first router that adds a label to an unlabeled packet. - Transit LSR: A core router that swaps labels. - Egress LSR: The last router that removes the label and delivers the IP packet. - LSP (Label Switched Path): The path that a labeled packet takes from ingress to egress. - FEC (Forwarding Equivalence Class): A group of packets that are forwarded the same way (e.g., same destination network). Each FEC is assigned a label. - Label: A 20-bit value in the MPLS header (32 bits total: 20-bit label, 3-bit EXP for QoS, 1-bit bottom of stack, 8-bit TTL).

To verify MPLS operation on a Cisco router:

show mpls ldp neighbor
show mpls ldp bindings
show mpls forwarding-table
show mpls interfaces

Example output for show mpls forwarding-table:

Local  Outgoing    Prefix            Bytes Label   Outgoing   Next Hop
Label  Label       or Tunnel Id      Switched      interface
16     Pop Label   10.1.1.0/24       0             Gi0/0      192.168.1.2
17     18          10.2.2.0/24       0             Gi0/1      192.168.2.2
18     No Label    192.168.3.0/24    0             Gi0/2      192.168.3.1

The Pop Label means this router is the egress LSR for that prefix.

Metro Ethernet (Metro-E) is a service that uses Ethernet technology to connect customer sites across a metropolitan area. It is often delivered over a service provider's MPLS backbone. Common Metro Ethernet services include: - E-Line: Point-to-point Ethernet connection (like a virtual leased line). - E-LAN: Multipoint-to-multipoint Ethernet LAN (like a VLAN across sites). - E-Tree: Rooted multipoint (one-to-many, e.g., hub-and-spoke).

The provider creates an EVC that maps to a specific VLAN ID. When a frame arrives at the PE from the CE, the PE adds a service VLAN tag (S-tag) to identify the EVC. The frame is then forwarded across the provider's MPLS backbone (often using MPLS over Ethernet) to the destination PE, which removes the S-tag and delivers the original frame to the destination CE. This is transparent to the customer, who sees a simple Ethernet connection.

Metro Ethernet services are often delivered over an MPLS backbone using technologies like VPLS (Virtual Private LAN Service) or VPWS (Virtual Private Wire Service). The provider uses MPLS labels to tunnel Ethernet frames across the core. This combines the simplicity of Ethernet with the scalability of MPLS.

In a typical enterprise network with multiple branch offices, MPLS VPNs are used to connect remote sites to a central data center. For example, a retail chain with 200 stores might contract a service provider to provide MPLS L3VPN connectivity. Each store has a CE router that connects to the provider's PE. The provider handles routing between sites using MPLS labels, allowing the enterprise to have a full-mesh or hub-and-spoke topology without managing their own WAN links. The network engineer at the enterprise only needs to configure a default route on the CE pointing to the PE, and the provider handles the rest. This is cost-effective and scalable.

Another scenario is a financial institution that requires low-latency connections between trading floors. They might use Metro Ethernet E-Line services, which provide dedicated bandwidth and deterministic performance. The provider guarantees a certain CIR (Committed Information Rate) and can prioritize traffic using EXP bits in the MPLS header. Misconfiguration can lead to traffic drops or SLA violations. For example, if the PE is not configured with the correct VLAN encapsulation, frames may be dropped or misdirected.

A common production issue is when the MPLS LDP session goes down due to a routing change or interface failure. This causes all labels to be removed, and traffic will be forwarded using regular IP routing (if the IGP still works) or dropped. Network engineers monitor LDP neighbor states using SNMP and set up alerts. They also use tools like 'ping mpls ipv4' to test LSP connectivity.