Cogent/Level 3 depeering (philosophical solution)
Patrick W. Gilmore
patrick at ianai.net
Sat Oct 8 15:10:26 UTC 2005
On Oct 8, 2005, at 7:02 AM, David Schwartz wrote:
>> Various people have stated that uneven data flows (e.g. from
>> mostly-content networks to mostly-eyeball networks) is a good
>> reason to
>> not peer.
>
> I think the industry simply needs to accept that it's more
> expensive to
> receive traffic than to send it.
But it is not.
It is more expensive to carry a large packet a long way than to carry
a small packet a long way. Because of things like hot-potato
routing, that frequently means the sender has less cost than the
receiver, depending on where they meet.
The rest of your argument is based on the premise that none of this
is changeable. Which is clearly wrong.
"Receivers" have been de-peering "Senders" for over half a decade.
(I.e. "Forever" in Internet time.) These fights have been fixed by
things like sending MEDs or intentionally recruiting customers to
balance traffic for a long, long time.
Of course, there is nothing wrong with an eyeball network saying
"I'll carry it, gimme gimme!" But that doesn't mean they have to.
> Yes, that can't possibly work. It's way too simple and actually
> makes
> sense.
No, it can't work because you assume things which are not necessarily
factual. Not to mention, it doesn't make sense.
--
TTFN,
patrick
More information about the NANOG
mailing list