On Thu, 8 Oct 2015, Fred Baker (fred) wrote:
I'm not sure why this discussion is happening on aqm@ instead of tcpm@... I
have added cpm to the cc line, and would recommend that anyone responding to
this thread do the same and remove aqm@.
I don't know why it started here, but not everyone on the aqm list is on the
tcpm list, so removing aqm@ will drop some people from the conversation.
On Oct 7, 2015, at 2:13 PM, David Lang <[email protected]> wrote:
So things that reduce the flow of acks can result in very real benefits to
users.
Dumb question of the month. What would it take to see wide deployment of RFC
5690? That would result in the data/ack ratio being reduced, on average, to
whatever amount had been negotiated.
well, given that people are still running Windows XP, it may get out to some
things but it's not going to get out to everything. Remember that VMware
responded to the EoL of XP with a campaign telling people they could keep
running XP in a VM.
in particular streaming media devices (smart TVs, etc) are not getting many, if
any, OS upgrades from the vendors, and they are a big part of the problem.
Summary for those that haven't read it - TCP implementations today generally
ack every other packet, with caveats for isolated or final data packets. This
proposal allows consenting adults to change that ratio, acking every third or
fourth packet, or every tenth.
If ack reductions are so very valuable, what's the chance of doing that on an
end to end basis instead of in the network?
a little less than the chance of shutting off IPv4 :-)
requiring changes to all the software on each end to enable this is wishful
thinking. The major servers could get the update pretty promtly, but updating
the client side?? not for a long time.
But in any case, that's addressing a slightly different issue. This discussion
started based on the fact that when there are queues, and the queues have
multiple acks for the same flow, it can save queue space, bandwidth, and reduce
latency by eliminating all but the last ack. If this can be done at a low cpu
cost (so it can be implemented on consumer endpoint hardware that usually has
the first bottleneck), it could be a win for the Internet as a whole.
David Lang
_______________________________________________
aqm mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/aqm
