Americas

  • United States
by Steve Taylor and Joanie Wexler

In the beginning…

Opinion
May 15, 20032 mins
Networking

* Frame perseveres but with formidable competitors

We’re continuing to take a look at what has changed – and what hasn’t – five years and 518 newsletters (but who’s counting?) since the first of our newsletters appeared in 1998 under the moniker “Frame Relay.”

Our first effort, “Still just a replacement for leased lines?” appeared on May 29, 1998.  We examined the tendency at that time for the topology of frame relay nets to mimic those of leased lines.  We noted that most networks still used a hub-and-spoke topology despite the then-popular concept of distributed computing with any-to-any connectivity.

Not much has changed.  Five years ago we said, “The reason these configurations persist is the same reason star-shaped T-1 networks were popular: It is too expensive to purchase separate circuits – virtual or physical – among many sites.”  And we’ll stand by this statement today.

Frame relay services continue to be priced on a model that mimics traditional leased-line pricing, deterring enterprises from using virtual circuit capabilities for any-to-any connectivity.  This has been disappointing. Only relatively recently have the services generically known as “IP-enabled frame” (and there are IP-enabled ATM subscriber services, too) started to break this pattern.

To be fair, in the mid-1990s, pioneering frame relay service provider WilTel (remember them?) launched a pricing plan that did provide incentive for meshed connectivity. WilTel dropped the cost of each additional virtual circuit you added with the idea that this would encourage more any-to-any virtual circuit topologies. However, this model proved too difficult to express in comparison pricing charts that appeared in the trade press at the time, and WilTel finally gave it up and returned to the more traditional port/PVC pricing.

Our second newsletter in May 1998, “Achieving a universal reach,” explored the problem of getting inexpensive any-to-any connectivity (IP) plus privacy and performance guarantees (frame relay) in a single service.  We pointed out – and it still holds – that even though Internet-based VPNs provide extensive reach, getting performance guarantees across a multicarrier network is problematic.  Nevertheless, there has been significant progress in the VPN space in this aspect, so frame relay now faces much more viable competition.