MPLS Explained - The Jargon You Need To Understand
MPLS - Multiprotocol Label Switching.
Multi-protocol - MPLS can be used to route data packets regardless of which protocols are being used, in the same way that Christmas wrapping paper (with labels!) can be applied to anything from a teddy bear to a PlayStation. So MPLS can be applied to Ethernet frames to allow a point-to-point layer 2 circuit to be set up over an MPLS network. MPLS can also be applied to IP traffic too.
Label - A 20 digit number made up of 0's and 1's. In addition to this label value, MPLS labels contain 3 bits for indicating a Class of Service tag, 1 bit to indicate that the label is the inner-most one (the one at the bottom of the stack), and 8 bits to denote a Time To Live value. MPLS networks often apply SEVERAL MPLS labels to a given packet. This enables them to separate different customers traffic AND reserve routes through their networks AND allow customers to prioritise their traffic within their available bandwidth.
WAN - Wide Area Network. Similar to the local area network (LAN) that connects the computers in your office, except the connections cover longer distances. WANs are typically used to link the LANs in various offices of the same organisation. MPLS is mainly used in service provider networks and in corporate WANs
Label Switch Router (LSR) - A router not at the edge of an MPLS network that looks at the outermost MPLS label and decides what to do with the packet based on that label.
Label Edge Router - MPLS routers at the edge of an MPLS network that do the heavy lifting adding labels when a packet enters the network. They remove any remaining labels if the packet is exiting the network.
Label Distribution Protocol (LDP) - Used by adjacent routers to tell each other about paths, so they can update their records and avoid sending traffic down paths that are dead ends.
VPLS - Not to be confused with MPLS. Virtual Private LAN Service. VPLS connections are often provisioned over MPLS networks.
Fast ReRoute - A feature of MPLS networks that makes it possible for traffic to be rapidly redirected to an alternative path (if one exists) in the event that the primary path becomes inaccessible. This cuts connection downtime from seconds to the blink of an eye.
