[150408] in North American Network Operators' Group

home help back first fref pref prev next nref lref last post

Re: IX in France

daemon@ATHENA.MIT.EDU (Jared Mauch)
Thu Feb 23 13:02:03 2012

From: Jared Mauch <jared@puck.nether.net>
In-Reply-To: <4F4679C7.8050102@gmail.com>
Date: Thu, 23 Feb 2012 13:00:51 -0500
To: virendra.rode@gmail.com
Cc: nanog@nanog.org
Errors-To: nanog-bounces+nanog.discuss=bloom-picayune.mit.edu@nanog.org


On Feb 23, 2012, at 12:39 PM, virendra rode wrote:

> I understand this is not true peering relationship, however its an
> interesting way to obtain exchange point routes and I understand this =
is
> nothing new.

<mini-rant>

I've found people who use the term 'peering' to mean something different =
than what I personally interpret it to mean.

eg: "We have peering with 4 carriers at our colocation facility where =
you can place gear"

Translation: We have blended IP transit from 4 carriers, or you can =
directly connect to them as needed.

I understand why they call it this, because "I configured peering with =
Level3/Cogent" on my router, etc.  The difference is in the policy.  =
What you're speaking of is someone selling transit, which is perfectly =
fine over various IXes, you generally are prohibited from 'selling =
next-hop', i.e.: you have to bear the cost on the IX port of the =
forwarding.

</mini-rant>

Buying transit isn't as dirty as people think it is, sometimes its the =
right business decision.  If you connect to an IX for $4000/mo at gig-e, =
you might as well buy transit at $4/meg on that same port IMHO.  You're =
unlikely to be using the port at 100% anyways at the IX, so your =
cost-per-meg there needs to properly reflect your 95% or whatnot.

- Jared=


home help back first fref pref prev next nref lref last post