Sicily to Egypt undersea cable disruption

Sean Donelan sean at donelan.com
Thu Jan 31 09:13:35 UTC 2008


On Thu, 31 Jan 2008, Martin Hannigan wrote:
>> From what I read about this cut, the way it happened seemed to have
> figurative odds of 1:1,000,000. It looks like "authorities" moved the
> anchorage area for some undefined reason. Cables are documented on
> marine charts and, at least theoretically under international
> standards, Captains and Pilots are lawfully required to refer to them
> before dropping the hook. Having some experience in marine operations,
> it would be 'curious' for a Captain or Pilot to not notice that there
> was a cable marking so close to their re-designated anchorage based on
> the chart that they would  need to  refer to for low tide depths and
> other (un)common hazards to insure that they weren't in imminent
> danger.

I'll leave the international law opinions to the lawyers rather than the 
network engineers :-)

> I'm sure that there is more to this story than meets the eye.

Single cable cuts are very interesting anymore because most networks have 
figured out most of those issues, usually by network darwinism.  Stuff
breaks "normally." There are the usual exception to the rule networks.

What makes this incident more interesting, as I indicated if its not 
one cable its another cable, was the double international cable cuts.
Likewise, what made Tawain 2006 interesting wasn't an earthquake affected
a cable, but there were multiple cable cuts in the region.

Quick, everyone get out your international cable maps and speculate
where in the world the next double (or triple, quad, etc) cable cut
could happen.  Due to regional politics, I don't think there are many
overland geographic diverse routes between countries to backup the
undersea routes.  If I remember the Wired article, FLAG did try to
build some overland geographic diversity through the region.

Stuff happens.  Although it will take a couple of weeks to repair these
cables (which seems to be the new "normal" repair time), I expect most
user traffic will be re-routed through less optimal but functional
routes within a few days.  Again, with the usual exception to the
rule networks.



More information about the NANOG mailing list