And your knowledge about fundamental networking rules has
demonstrated
a significant disparity from the knowledge of quite a few other
folks. Any literature references to this fundamental rule would
be appreciated.
The effects of layered multiplexing were mostly experienced when
trying
to run stuff TCP/IP on variation of protocols doing their own flow
control
Interesting. CSNet (the predecessor to NSFNet) ran TCP over X.25 just
fine. (There are lots of other examples, but that one rather stands
out.)
Well, I remember running TCP over the Transpac's X.25 network at about
the same time as CSNET, and my memories would certainly not qualify as
"fine." What did happen quite frequently was that one of the many TCP
connections would send a burst of packets that would saturate the X.25
flow control; this would result in long delays, leading to timers on the
other VCs. The underlying X.25 network, Transpac, may well have behaved
differently than CSNET; the transmission speed were about 9600 bps; so
that may explain the differences. Also, this was before the introduction
of congestion control and the revision of the RTT estimation algorithm.
But I can certainly see variations of the problem happening in BEEP,
e.g. if the buffer estimation at the BEEP layer goes out of whack with
the actual buffer management at the TCP level.
-- Christian Huitema