Why the Mediterranean Is the Net's Achilles' Heel

An anonymous reader writes "A spate of broken cables has brought disruption for many of the world's Web users in 2008 — and the Med has been at the center of the problems. For political reasons, the Mediterranean Sea is an Internet bottleneck through which the majority of traffic between Europe and Asia is squeezed. That traffic must run the gauntlet of earthquakes and heavy maritime traffic to reach its destination. Better and stronger cables are urgently needed to avoid a re-occurrence of the 2008 outages."

Comments

Jeez.

[girlintraining]

In the 90s it was backhoes. Now it's giant cable-eating squid. What next, volcanic eruptions? Really, the problem is two-fold -- first, cables break. Hey, it's several thousand miles long and several thousand feet down, and it's just laying there. Of course it's going to break. You could make the cables out of Unobtainium and they will still wither and break eventually. It's a fact of life. The real problem isn't that they fail, the problem is that the telecommunications companies don't have redundant links because of the expense. So, in summary, the problem is economics. And Cthulu. But you can't stop one of the great old ones, so let's focus on redundant links instead. -_-

Re:Jeez.

[Red Flayer]

So, in summary, the problem is economics. And Cthulu.

He Who Lies Dead but Dreaming has no part to play in the damage to undersea cables, I have this on good authority. The Telcos are actually agents of Cthulhu (duh! -- you should know this by now if you've ever called telco tech support); the internet is just one of his dreams, which will serve to increase chaos and drive us all to madness.

Seriously, though, blaming the problem on economics is a copout. Why are costs to lay redundant cables so high? What can be done to convince the telcos that laying redundant cables is a good idea? What can tip the CBA to the B side?
(br>There are lots of reasons a truly redundant system is prohibitively expense. The cost of negotiating rights-of-way through multiple nations, for example. The increased costs to shipping (external cost to the telcos) from avoiding cable paths (and this is magnified with true redundancy, since redundant cables should not follow the same path). The costs of running and maintaining landlines in politically unstable areas. And, not least of all, the costs in materials, capital, and labor to run redundant lines.

The way to tip the scale in favor of running redundant lines is to either reduce the cost of doing so, or increase the benefit from doing so. How much money do the telcos lose when a line goes down? Over time, is that more than the cost of running redundant lines?

So yes, it's economics, but saying it's economics is glossing over the important details.

Overstated consequences

[ninti]

The article seems a little alarmist. For instance, this line: "The 2008 outages hit local economies hard and a stronger quake could plausibly bring Mediterranean economies to their knees, by denying them access to crucial global markets for days or weeks. A 2005 study at the Swiss Federal Institute of Technology Zurich calculated that a nationwide internet blackout would cost Switzerland 1% of its GDP per week." But of course a cut in the Mediterranean will not be a "nationwide internet blackout" for Switzerland much at all. In fact, if India and the mid-east gets cut off from the rest of the Internet, the rest of the world won't care all that much.