On Wed, Jun 22, 2011 at 12:10 AM, Rick Jones <[email protected]> wrote: > Perhaps, though at 100 ms RTT, which I would consider non-trivial > latency :) (*) one does indeed need something like 1.8 MB of TCP > window to get link-rate for an OC-3. Â Montgommery Scott may have the > ability to appear to change the laws of physics, but I don't think > anyone else can :)
> (*) assuming, of course that the 100ms wasn't queing delays and was > indeed speed-of-photon. Some was speed-of-light, some was queuing. Since the Earth has a diameter of about 25K miles and light does 186K miles/s, in theory, the worst case light speed RTT for Earth communications is going to be about 130ms. In practice, you don't get anywhere near lightspeed because the link cables aren't following the ideal path between source and dest; network gear adds overhead; and your link path may bounce a signal off a high-latency satellite link instead of always being able to use cables. I've had legit RTT as high as 600ms, but that was on a much slower link, so TCP actually required no tuning at all to max out that circuit. - m _______________________________________________ questions mailing list [email protected] http://lists.ntp.org/listinfo/questions
