"Luke S. Crawford" <lsc@prgmr.com> writes:
On Sat, Apr 07, 2012 at 06:16:30PM -0400, Robert E. Seastrom wrote:
Sometimes making the AS path as short as possible makes a lot of sense (e.g. when trying to get an anycast network to do the right thing), but assumptions that peering results in lower costs are less true every day.
I keep reading people say that. But wouldn't the same forces that push down the per-megabit cost of transit also push down the per-megabit cost of peering?
Generally the costs of transit are pushed down by competition. As a vendor your costs for bandwidth/transport/port*bw may drop but you are unlikely to drop your prices to your customers merely because your costs have gone down unless prompted to by a competitor. In any given IX, cross-connect fibers and peering switch ports are often a monopoly. While not unheard-of for there to be two competing IX switch fabrics available in a single facility, the cross-connects to those competing exchanges are not free, and I'm not aware of any sizeabe facilities that are still "run your own XC and don't pay anyone for it" (of course, as soon as I say that I'll get private email or an IRC message pointing out the corner case). Consider the case of a peering n00b network (the target of this discussion after all) in hypothetical facility that charges $1000/month for a gigabit ethernet port on the peering fabric. You turn up a connection to this port and discover that (without buying people drinks / sushi dinners / etc at a conference) you can bring up enough peering with other networks to move 150 Mbit/sec on it. That's pretty optimistic for a small player, but still... now you're paying $6.66/mbit for that transit. If you can move 150 Mbit/sec to low-hanging-fruit transit you're probably between 1 and 2gbps total. How's that compare with what you're paying for transit with that level of commit? -r