Internet Exchange Visualization
John van Oppen
john at vanoppen.com
Mon Aug 28 15:29:01 UTC 2023
I've always been a little less harsh than What Jared mentions, but my theory is like within say 5-7 ms is probably reasonable as long as the endpoint is closer than the next major IX both are present on. I don't really know what folks think they are getting by peering across the world. I think this might be one of those vanity peering type situations instead of any real technical justification. I have a hard time understanding how it would not often make routing worse.
John
More information about the NANOG
mailing list