In 1992, frame relay was portrayed as an interim technology that would be replaced by Switched Multimegabit Data Service. In 1995, industry insiders predicted frame relay would be devoured by ATM. Today, many believe frame relay will fade away as everyone moves to IP-based VPNs. But don’t hold your breath -- frame relay will be around for a long time.

There are several keys to frame relay’s longevity. One advantage is its simplicity. Frame relay is, in the best sense of the word, a dumb protocol. There’s no extra complexity, and implementations are simple. Making frame-relay-compliant premise equipment is easy, and any complexity, such as adding differentiated classes of service, is performed within the network.

The frame relay protocol is also efficient. Its header is the same size as that of PPP, typically used as an alternative with IP. So there’s no penalty for using frame relay as the Layer 2 protocol for IP traffic. And for non-IP traffic, there’s no need to encapsulate in IP.

Frame relay pricing is also simple. The elements on which pricing is based -- access plus ports for physical connectivity pricing plus a per-permanent-virtual-circuit price based on the committed information rate -- make comparison with dedicated services easy. In typical configurations, using frame relay results in savings of at least 50% compared with dedicated services that have comparable performance.

The connection-oriented nature of frame relay also contributes to its longevity. With the capability to support roughly 1,000 virtual circuits per frame relay User-to-Network Interface even without extended addressing, frame relay can support most connectivity needs. The connection-oriented architecture also provides an intrinsic level of security because there’s no opportunity for address spoofing: Network addressing is controlled within the service provider’s network. The connection-oriented commonality with ATM services makes interworking between these services a breeze.

Finally, users’ reluctance to change from something that works bodes well for frame relay. If you can save money using a tried-and-true technology, there is little incentive to change.

Of course, there are challenges service providers must meet to give frame relay an even brighter future. More differentiated classes of service for supporting traffic types are needed, as are more "IP-aware" services that take advantage of optimizing frame relay for transporting IP traffic. But perhaps the biggest test is the pricing challenge from Internet-based services. While the ‘Net can’t deliver the quality of service and dependability that frame relay services do, there’s a limit to the pricing differences that can be justified. Frame relay prices must drop to meet this challenge.

But if you’re feeling a bit behind because you haven’t dumped your frame relay network for something new, take heart. You’re in good company.

» posted by ITworld staff

Network World

• Email
• Print
• Share
  • Slashdot
  • Digg
  • Stumble
  • Reddit
  • Newsvine