[31620] in North American Network Operators' Group
Re: Statistical Games Providers Play (RE: availability and
daemon@ATHENA.MIT.EDU (Sean Donelan)
Sun Oct 1 03:13:58 2000
Date: 1 Oct 2000 00:11:18 -0700
Message-ID: <20001001071118.27710.cpmta@c004.sfo.cp.net>
Content-Type: text/plain
Content-Disposition: inline
Mime-Version: 1.0
To: vijay@umbc.edu
From: Sean Donelan <sean@donelan.com>
Cc: nanog@merit.edu
Errors-To: owner-nanog-outgoing@merit.edu
On Sat, 30 September 2000, Vijay Gill wrote:
> There are a lot more variants regarding the routing architecture (IGP
> setup, bgp setup, et al), and depending on various failure modes, some are
> better than others for a subset of failures and vice versa.
True, but in the end does it end up being a zero-sum game? Or are there
real differences in performance? I'll pick on a couple of different providers,
but we could use anyone.
In my experience AT&T has a huge MTBF, over 7 years when I bought circuits.
But when the two natural disasters struck at the same time, it would take
AT&T several days to get the circuits working again. On the other hand,
Sprint had a problem every month or two, but they usually had them fixed
in about 20 minutes. What's the trade-off. Over 10 years, the availability
numbers weren't that different between AT&T and Sprint.
Sprint hypes their SONET fiber network, AT&T hypes their FASTAR network
restoration. Is it strictly a question of cost? Although a lot of
advertising and sales emphasis is placed on the technology, I haven't
found the technical differences between providers affecting the delivered
performance. Non-technical factors seem to have a bigger affect.
