Virtual private local area network services (VPLS) have long been in use by an organization needing prime-time availability for an international wide area network or metro area network. VPLS puts Ethernet connections in point-to-point, multipoint, or multi-point to multipoint configurations. This framework for a geographically dispersed or global organization is a welcome architecture. One of the other popular alternatives is to use Internet-based VPN connections, which offer different reliability than a VPLS connection. AT&T recently announced a new offering of VPLS connections that are available in speeds under 1o Mbps and available in points around the world. Having the speeds available under 10 Mbps saves network infrastructure costs in not having to replace an existing copper infrastructure to go to fiber connectivity at a VPLS endpoint.
One distinguishing factor that sets VPLS connections apart is that the connections are on the same network when provided by the same carrier. In this situation the remote locations are on a single network from a TCP/IP perspective, and may be easier to manage from a top level network management perspective. This would include managing a redundant path of a network that could use an entirely different source. For example, let us take the scenario where multiple sites are connected by Internet-based VPN and a VPLS endpoint is available at each site. In this example the Internet VPN would be the primary mechanism for intra-site communication, and the VPLS is for certain traffic that from a regulatory or compliance reason cannot go over the Internet. The VPLS connection can also serve as a secondary network in the event that the Internet connection becomes unavailable at a remote site or as a channel for a critical business or trading partner as an extranet. Because the VPLS endpoints are on the same network, the rerouting of a site's Internet connection would be fairly straightforward with another persistent connection available.