Multiprotocol Label Switching (MPLS) is growing in popularity as a WAN technology used by service providers. This is due in part to MPLS’s capability to support multiple protocols on the same network. MPLS also has the capability to perform traffic engineering (which allows traffic to be dynamically routed within an MPLS cloud based on current load conditions of specific links and availability of alternative paths).
MPLS inserts a 32-bit header between Layer 2 and Layer 3 headers. Because this header is shimmed between the Layer 2 and Layer 3 headers, it is sometimes referred to as a shim header. Also, because the MPLS header resides between the Layer 2 and Layer 3 headers, MPLS is considered to be a Layer 2 1/2 technology.
The 32-bit header contains a 20-bit label. This label is used to make forwarding decisions within an MPLS cloud. Therefore, the process of routing MPLS frames through an MPLS cloud is commonly referred to as label switching.
An MPLS frame does not maintain the same label throughout the MPLS cloud. Instead, an LSR receives a frame, examines the label on the frame, makes a forwarding decision based on the label, places a new label on the frame, and forwards the frame to the next LSR. This process of label switching is more efficient than routing based on Layer 3 IP addresses. The customer using a provider’s network and the MPLS transport across that network is not normally aware of the details of the exact MPLS forwarding that is done by the service provider.