[prev in list] [next in list] [prev in thread] [next in thread] 

List:       inet-access
Subject:    Re: InterNAP
From:       Ben Black <black () layer8 ! net>
Date:       1999-04-30 23:48:31
[Download RAW message or body]

On Fri, Apr 23, 1999 at 03:31:02PM -0400, Avi Freedman wrote:
> In article <99320.174242.14453@avi.netaxs.com> you wrote:
> 
> : As for peering (the sort that occurs at existing public NAPs), it is
> : not really appropriate in the InterNAP model.  As the bandwidth
> : wars between GTEi and Exodus (among others) illustrate, bandwidth
> : exchange without per-bit settlement can get very messy.  You avoid
> : it by paying for what you use.  "Peering" with InterNAP at a given
> : PNAP is handled in exactly the same fashion as the rest of the 
> : connectivity to the PNAP: by purchasing bandwidth. 
> 
> I'd argue that it's appropriate in ANY model.
> 
> InterNAP customers will get demonstrably better connectivity to, say,
> nac.net or lightning.net over a private 100bt than via one of their
> transit providers, no matter who that provider is.
> 
> The thing that is scary about InterNAP is that with so many transit
> providers out of each city, all it takes is any one of 9 or however
> many transit providers to screw up and you lose connectivity to some
> sites, or have loss or what have you.
> 

I'm not clear on how you figured this one out.  Is it not the case that
if an AboveNet (or any other provider) link used for peering fails, you
may lose connectivity to some sites or have loss?  Using local prefs
does not cause BGP to suddenly stop using perfectly good routes.

-- 
 --b

This time my name is Ben.

-
Send 'unsubscribe' in the body to 'list-request@inet-access.net' to leave.
Eat sushi frequently.   inet@inet-access.net is the human contact address.

[prev in list] [next in list] [prev in thread] [next in thread]