Re: [dccp] dccp spec expert review (Minshall, main spec)

[Greg Minshall <minshall@acm.org>]

hi, Eddie,

> A notices the drop and halves its congestion window, to 4,150 packets.
> The sustainable congestion window now is 7,916 packets. It will take
> 3,405 round-trip times -- more than 5 minutes -- to get back to the
> sustainable cwnd. 

i don't find the example convincing.  what if brief congestion "inside the 
network" or a line hit had caused a packet to be lost?  i assume in either of 
these two cases it will take 5 minutes to get back to "full speed"?

(in terms of "data dropped" for corruption losses, i doubt most corruption 
losses will be detectable at the receiving end host -- either because they 
will have occurred before reaching the end host, or will have mangled the 
packet to the extent that the end host can't determine the specific 
connection.  thus, i don't think "data dropped" is a realistic way of 
distinguishing between congestion and corruption losses.)

cheers, Greg


_______________________________________________
dccp IETF mailing list: dccp@ietf.org
list info:  https://www1.ietf.org/mailman/listinfo/dccp
wg charter: http://www.ietf.org/html.charters/dccp-charter.html