[72225] in North American Network Operators' Group
Re: Peering point speed publicly available?
daemon@ATHENA.MIT.EDU (bmanning@vacation.karoshi.com)
Fri Jul 2 22:14:19 2004
Date: Sat, 3 Jul 2004 02:13:35 +0000
From: bmanning@vacation.karoshi.com
To: Erik Amundson <erik@myevilempire.net>
Cc: nanog@merit.edu, network.support@oati.net
In-Reply-To: <20040702000943.6572CCC0EC@mail.myevilempire.net>
Errors-To: owner-nanog-outgoing@merit.edu
On Thu, Jul 01, 2004 at 07:09:52PM -0500, Erik Amundson wrote:
> NANOG,
> I have a question regarding information on my ISP's peering relationships.
> Are the speeds of some or all peering relationships public knowledge, and if
> so, where can I find this? By speed, I mean bandwidth (DS3, OC3, 100Mbps,
> 1Gbps, etc.). I am trying to transfer large stuff from my AS, through my
> ISP, through another ISP, to another AS, and I'm wondering how fast the
> peering point is between the ISPs. I'm working with my provider to get this
> information as we speak, but I'm wondering if it's available publicly
> anywhere. If it were, this could be one way to evaluate providers in the
> future, I guess.
perhaps you have already beaten this dead horse enough but
here is my non-flash, stick/ascii rendition...
{ ISP core } --- [rtr] --- [possibleIX] --- [rtr] --- { ISP core }
lkA lkB lkC lkD
lkA & lkD are "hidden" from external view. if you are
a customer of ISP which owns lkA, they -may- tell you
what the characteristics of lkA are ... "today". No
assurances that it will remain the same for any given
period of time. And the ISP with lkA is unlikely to be able
to judge/interpret the accuate value of lkD, although this
may be infered from their peering SLAs... which are generally
NDA'ed.
if ther exists a "possibleIX", there is the strong case
that the interconnects are one of the possible ehternet
formats, e.g. 10Mbps in full or half duplex, 100Mbps in
either full or hald duplex, or 1Gbps, generall full duplex.
Yes, there are some other varients... :) And there is no
assurance that lkB and lkC are the same!
in the case where there is no "possibleIX" and the links
lkB and lkC are the two sides of a point2point link, then
more choices arise and the potential for determining
the actual "speed" of the link.
then there are the other "tweekable" characteristics of
a specific link (MTU, MSS, etc.) which will affect throughput/goodput
of that specific link.
Sorry for the additional flogging. :)
--bill (wading thorugh two weeks of old email)