[72224] in North American Network Operators' Group
Re: Peering point speed publicly available?
daemon@ATHENA.MIT.EDU (Stewart, William C (Bill), RTSLS)
Fri Jul 2 21:32:04 2004
Date: Fri, 2 Jul 2004 20:31:16 -0500
From: "Stewart, William C (Bill), RTSLS" <billstewart@att.com>
To: <nanog@merit.edu>, <erik.amundson@oati.net>,
<erik@myevilempire.net>
Errors-To: owner-nanog-outgoing@merit.edu
On or about July 1 2004, Erik <allegedly at> myevilempire.net> Amundson=20
allegedly asked about peering point bandwidth.
Some North American ISPs will tell you that under non-disclosure,
but almost all of them will point you to their standards for peering,
and you won't find many Tier 1 ISPs that peer at less than DS3 in the =
US,
and probably not many in Canada, London, Amsterdam, Tokyo, or Singapore =
either.
That means the insertion delay is under 0.27ms, or about 27 fiber-miles,
so it's less important than whether the peering is in San Francisco or =
San Jose.
Queuing due to overload is really much more important than absolute =
size.
Also, if you're dealing with ISPs that use public peering points,
those may be a performance concern, but in the US that's mostly not =
Tier1-Tier1.
(Linx is a different case entirely, assuming you want your traffic to be =
in London.)
Smaller ISPs might be more talkative, or if you're having actual =
problems,
like why your connection from minneapolis.example1.net to =
stpaul.example2.net=20
goes through a peering point in San Francisco instead of peering in =
Minnesota
or at most Chicago, big ISPs can also be pretty talkative. =20
