[e2e] TCP, retransmissions, and ISPs with byte caps

[RJ Atkinson]

On Tuesday, June 18, 2002, at 08:57 , David P. Reed wrote:
> At 08:17 AM 6/18/2002 +0200, Philippe Strauss wrote:
>> But about delays, CATV systems have an interesting characeristic:
>> a ~20ms minimum delay in the ACK path (direction from home
>> to the hub, upstream or return path in catv speak).
>
> I have been pinging various targets through my LANCity cable modem for 
> years now and I have regularly gotten 7 msec. ping RTT to lcs.mit.edu 
> which is a few hops (11 at the moment) away.  This morning the average 
> is 13 ms.   Where's the 20ms minimum?   How did you determine it?

Of course latency varies with the technology used (DVB, LANcity, CDLP, 
DOCSIS, etc)
and with the actual distance as the copper/fibre runs from CM to CMTS.

And LANcity is an interesting case.  In LANcity the upstream is the
same as the downstream in LANcity (half-duplex, symmetric, 10 Mbps,
and coupled-together).  This is a nice counter-example to Phillipe's
assertion that the upstream is always different than the downstream.

Ran