[97564] in North American Network Operators' Group
Re: TransAtlantic Cable Break
daemon@ATHENA.MIT.EDU (Frank Coluccio)
Sat Jun 23 00:05:04 2007
From: Frank Coluccio <frank@dticonsulting.com>
To: Hank Nussbacher <hank@efes.iucc.ac.il>,
Sean Donelan <sean@donelan.com>
Reply-To: frank@dticonsulting.com
Date: Fri, 22 Jun 2007 22:59:29 -0500
Cc: nanog <nanog@merit.edu>
Errors-To: owner-nanog@merit.edu
Interestingly, some major transoceanic undertakings have begun looking very
favorably towards a meshed topology solution, eschewing rings. Verizon is
championing this approach at the present time as a consortium partner in the
Trans-Pacific Express (TPE) cable laying venture to China, and offers
justifications for doing so in several interviews and PPT prezos I've come
across. Makes sense. From: http://preview.tinyurl.com/yqcrzm
Article "Verizon Business plots business plan for =E2=80=9907 and beyond"
May 2007 Issue of Lightwave Magazine
<snip>
"Last year, Verizon also took steps to improve the reliability of the
transatlantic portion of its global IP network. In the past, traffic moved =
across
the Atlantic over SONET rings, which provided redundant paths. However, such
architecture only protects against a single failure in a given ring. A fail=
ure or
service interruption on two or more segments of the same network required t=
he
deployment of a cable ship-the nautical equivalent of the truck roll-to res=
tore
service. Today, Verizon Business operates a mesh network, using Ciena
CoreDirectors (www.ciena.com), to move traffic between six diverse paths th=
at can
be routed onto other undersea networks in the event of a network failure."
<snip>
Of course, these offerings are still being "productized," so it remains to =
be
seen what terms and conditions they carry, and how they will be priced.=20
Frank
=3D=3D
On Fri Jun 22 11:56 , Sean Donelan sent:
>
>On Fri, 22 Jun 2007, Hank Nussbacher wrote:
>>> Tell that to the 10 gig wave customers who lost service. Very few cable
>>> systems provide protection at the 10 gig wave level.
>>
>> If you don't pay the extra amount for a protected circuit, why should yo=
ur=20
>> circuit get protection for free when others have to pay for it? Now, if=
=20
>> there are 10G customers with protected circuits who lost service, then=
=20
>> hopefully they have in their contract hefty penalty clauses against the=
=20
>> carrier. If not, then they are just plain stupid.
>
>Is paying for "protected circuits" actually worth it. Or are you better=
=20
>off just buying two circuits and using both during normal conditions.=20
>Use switching at layer 3 to the remaining circuit during abnormal=20
>conditions. Most of the time, you get twice the capacity for only twice
>the price instead of a "protected circuit" where you only get the once=20
>the capacity for twice the price.
>
>Of course, there is still the problem some facility provider will "groom"=
=20
>both your circuits on to the same cable. If you are buying pre-emptable=
=20
>circuits, hopefully you understand what that means.
>
>
>
