Latency and Distance

Generally, latency, or time delay, is based on speed and distance; traveling at twice the speed, or for only half the distance, can halve the latency. However, there are a number of complicating factors in the real world, for both speed and distance.

There is—at least with current physics—a limit to speed: the speed of light in a vacuum. This theoretical speed is not actual speed: Light in a fiber travels significantly more slowly than light in a vacuum, and traffic congestion impacts real speed.

Another issue is that the distance as the crow flies is not necessarily the same distance that a routed network covers. There are all kinds of anomalies in routed networks due to cable routes that follow everything from rights-of-way on train tracks and bridges to terrain topology constraints due to mountains, valleys, rivers, lakes, and coastlines to where cities happen to have sprung up—often due to those same constraints. Undersea, there are topological anomalies to be managed, areas of seismic activity requiring wide berth, and fishing areas—where cables tend to be dug up—to be avoided.

TeleGeography, a telecom market research firm covering international networks, was kind enough to provide Exhibit 19.1, a map of undersea cable routes. Although the routes don’t exactly follow the paths shown due to some liberty taken with graphics for readability, a quick glance at the map shows the uneven deployment of routed networks. In some cases—say, New York to London—there ...

Get Cloudonomics: The Business Value of Cloud Computing, + Website now with the O’Reilly learning platform.

O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.