For the history of the telecommunications industry, there have been some schisms within the networking organization. One of the most fundamental was always between the “WAN” folks and the “LAN” folks. And there were good reasons.
LAN & WAN equipment vendors feel the crunch
Let’s start with bandwidth availability. Within the LAN, it’s always been a pretty good assumption that bandwidth is basically free and unlimited. In the WAN, bandwidth has been assumed to be expensive and limited.
Consequently, with a fundamental difference that’s this profound, it’s only logical that the data protocols would reflect these underlying assumptions.
Ethernet is the perfect example of a transport protocol that allows for excellent interconnectivity with little (if any) configuration needed. On the flip side, it’s extremely overhead intensive in that each payload of information takes a lot of additional information wrapped around it to ensure that it is appropriately delivered.
On the WAN side, the luxury of an overhead-intensive protocol was never an option. Frame relay, for example, was a great success because it is a very “skinny” protocol that delivers tremendous bandwidth efficiency. However, from a “reachability” perspective, it’s hardly more than a private line. Other equipment must take the information delivered by frame relay and do the more overhead-intensive heavy lifting of ensuring connectivity.
As mentioned in the last newsletter, we may very well be reaching a major inflection point in the way we think about networks. In particular, the labor-intensity of maintaining multiple systems and the rapid decline in the drop of WAN bandwidth (in terms of cost per bit per second) seem to be changing the fundamental ground rules for WAN networking, so that a LAN-like “free and unlimited” assumption may not be totally incorrect.
Still, this does bring up a wide range of questions that have to be answered. For starters, by moving to a WAN Ethernet service, the “WAN folks” are giving up a lot of control to the service provider. And this is a level of trust that the service providers are still earning.
Further, the range of technical issues questions like latency, jitter, and other key performance parameters are being “trusted” to the network provider. Are you ready for that?
And then there’s pricing. For starter’s what’s included in a “Committed Data Rate”? How is this enforced? What types of controls are available for this.
For starters, we’ve opened up a thread on this topic at TECHNOtorials, and we invite your participation in the discussion.
We’re going to take a short break from this topic, but we’ll be back soon, and we’re hopeful that when we return we can point you toward a “showdown” where the major suppliers can tackle these questions head-on both with each other and with potential customers.
