RE: TCP (and SCTP) sucks on high speed networks

["Dr. Carsten Bormann" <cabo@tzi.org>]

> TCP's "congestion avoidance" algorithms are not compatible with
> high speed,
> long distance networks.  The "cut transmit rate in half on packet loss and
> increase the rate additively" algorithm will simply not work.

I don't believe this is just a matter of algorithm.
The problem is really the dynamic range of the rate adaptation equation.

> Consider a 10Gbs link to a destination half way around the world.
>  A packet
> drop due to link errors (not congestion or infrastructure products) can be
> expected about every 20 seconds.  However, with a RTT of 100ms (not even
> across the continent), if a TCP connection is operating at 10Gbs,
> the packet
> drop (due to link error) will drop the rate to 5Gbs.  It will
> take 4 *MINUTES*
> for TCP to ramp back up to 10Gbps.

If you want to stay in the regime of the TCP equation, 10 Gbit/s for a
single flow of 1460 byte segments at 100 ms (quarter-earth) RTT means you
need a packet loss rate of about 1E-10, i.e. you should lose packets about
once every five hours.
Clearly, this kind of congestion signal is way too coarse for any useful
control algorithm.

Gruesse, Carsten