Long hops on international paths

William Herrin bill at herrin.us
Tue Jan 18 20:44:26 UTC 2022


On Tue, Jan 18, 2022 at 6:49 AM PAUL R BARFORD <pb at cs.wisc.edu> wrote:
> So, the question is what is the cost/benefit to providers to
> configure/maintain routes (that include long MPLS tunnels)
> that tend to concentrate international connectivity at a
> relatively small number of routers?

Most likely that's not what's happening. None of your traces showed
traffic going to Chicago and then doubling back. I can't say with
certainty what this means but my educated guess is this:

Chicago is a major peering point for the networks in question, so
packets headed more or less that direction tend to hop over there
instead of somewhere else. Having crossed over, they enter the MPLS
system which is opaque to traceroute, so they're not seen again until
they exit MPLS. Traffic originating closer to the final destination
goes to a scattering of other MPLS entry points instead, so Chicago
looks outsized in the data. The actual traffic flow, opaque inside
MPLS, goes through the nodes you'd expect, close to the cable
landings.

So it's not a question of whether the traffic has an international
destination, it's a question of whether that backbone is in the path
to the destination (international or not). Rephrase your question as:
is there a cost/benefit to providers having a small number of very
large peering points accepting traffic into their networks in addition
to the myriad small ones.

Regards,
Bill Herrin


--
William Herrin
bill at herrin.us
https://bill.herrin.us/


More information about the NANOG mailing list