<div dir="ltr">I'm definitely interested in seeing if the new pie implementation fares better than what I was seeing on 3.10.24-8.<div><br></div><div>-Aaron</div></div><div class="gmail_extra"><br><br><div class="gmail_quote">
On Mon, Jan 20, 2014 at 4:46 PM, Dave Taht <span dir="ltr"><<a href="mailto:dave.taht@gmail.com" target="_blank">dave.taht@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<a href="http://www.spinics.net/lists/netdev/msg264935.html" target="_blank">http://www.spinics.net/lists/netdev/msg264935.html</a><br>
<br>
Hat off to vijay and the pie folk at cisco who shepherded the code<br>
through 5 releases to get it upstream!<br>
<br>
PIE is now mandated in DOCSIS3.1 cablemodems and the ECO for DOCSIS<br>
3.0 went out a while ago.<br>
<br>
I think the mandate differs from the Linux implementation. I don't<br>
know the size of the recomended cablelabs target nor the size of the<br>
estimation window (?)).<br>
<br>
So that will improve going-upstream buffering problems dramatically.<br>
<br>
The linux implementation supports ECN, the DOCSIS standard does not.<br>
The ECN work in pie has some minor problems that fq_codel doesn't<br>
have, we hope to discuss at ietf...<br>
<br>
I will fold the v5 implementation of pie into cerowrt in the next release and<br>
see what happens.<br>
<br>
NOW:<br>
<br>
The biggest problem remaining in cable is fixing excessive downstream<br>
buffering. Surely something can be done to remove the 1.6 seconds of<br>
buffering I observed yesterday (at some level without needing to AQM?<br>
Some workaround that operators can use?). Some benchmarks I ran last<br>
night (not against pie):<br>
<br>
<a href="http://snapon.lab.bufferbloat.net/~d/bev/comcast_native_ipv6/" target="_blank">http://snapon.lab.bufferbloat.net/~d/bev/comcast_native_ipv6/</a><br>
<br>
I sure hope the gpon, lte, and wifi folk are paying attention.<br>
<span class="HOEnZb"><font color="#888888"><br>
--<br>
Dave Täht<br>
<br>
Fixing bufferbloat with cerowrt: <a href="http://www.teklibre.com/cerowrt/subscribe.html" target="_blank">http://www.teklibre.com/cerowrt/subscribe.html</a><br>
_______________________________________________<br>
Bloat mailing list<br>
<a href="mailto:Bloat@lists.bufferbloat.net">Bloat@lists.bufferbloat.net</a><br>
<a href="https://lists.bufferbloat.net/listinfo/bloat" target="_blank">https://lists.bufferbloat.net/listinfo/bloat</a><br>
</font></span></blockquote></div><br></div>