On Mon, 12 Oct 2015, Jonathan Morton wrote:
My contention is that since this is already happening, consolidating the
ACK packets into stretched ACKs doesn't make this any worse, and it saves
network bandwidth (and decreases latency to the extent that data is acked
faster than waiting for the entire chain or original ACKs to get through,
especially if that would take multiple transmit windows). As a result,
thinning the ACKs is kinder to the network.
I agree, *iff* they are not DupACKs signalling packet loss. Do existing
and future cable modems take that subtle distinction into account?
I don't know, and RFC3449 section 5.2.1 lists a couple other conditions to
watch out for. But this discussion started with what AQM algorithms can/should
do when they see multiple ACKs in their queue (and I added wifi as part of
make-wifi-fast). So we are talking about what should be put in future code. The
cable modems are just an example we can draw from of a past implementation of
similar ideas.
hopefully future cable modems will use a more generic AQM algorithm instead of
the hard-coded optimizations they've had in the past. It's almost always a win
when a generic solution can replace a specific solution :-)
David Lang
_______________________________________________
aqm mailing list
aqm@ietf.org
https://www.ietf.org/mailman/listinfo/aqm