<div dir="ltr">Here's a long-range (indoors) test result with the 1900AC. I have about 9dB snr at the client (all the way across my house, which has plaster walls):<div><br></div><div><a href="http://www.dslreports.com/speedtest/737736">http://www.dslreports.com/speedtest/737736</a><br></div><div><br></div><div>download buffering is the 1900ac, upload buffering is the mac laptop's driver. Both are trying too hard to deliver those packets, I think.</div><div><br></div><div>The driver has crazy over-buffering in the face of poor signal conditions. OTOH, I'm amazed at the throughput numbers. But this is an 80Mhz wide 5Ghz channel, in a quiet environment, so it has LOTS of airspace to work in.</div><div><br></div><div>-Aaron</div></div><div class="gmail_extra"><br><div class="gmail_quote">On Wed, Jun 24, 2015 at 9:32 AM, 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"><span class="">On Wed, Jun 24, 2015 at 4:31 AM, Mikael Abrahamsson <<a href="mailto:swmike@swm.pp.se">swmike@swm.pp.se</a>> wrote:<br>
> On Tue, 23 Jun 2015, Sebastian Moeller wrote:<br>
><br>
>> As Dave said it would be nice see RRUL data from the same testbed. It<br>
>> would be so nice if flint had a way to send different sized TCP packets… (I<br>
>> guess this might be faked with MSS clamping in the router and relaying on<br>
>> path MTU discovery?)<br>
><br>
><br>
> What kind of tests should I run? I have rrul results now, both at the same<br>
> time as running iperf3. I set iperf3 to run with 10 parallel streams,<br>
> different MSS per test, and let it run for 30 seconds into the rrul test. So<br>
> in all the tests, iperf3 session stops running half way into the rrul test.<br>
><br>
> I set sqm to 500M up and down on eth0, ECN up and down, and not to squash<br>
> DSCP in any direction.<br>
><br>
> <a href="http://swm.pp.se/aqm/rrul-wrt1200ac-150624.pdf" rel="noreferrer" target="_blank">http://swm.pp.se/aqm/rrul-wrt1200ac-150624.pdf</a><br>
<br>
</span>From what I see here you are rarely, if ever, engaging fq_codel<br>
properly. Latencies are pretty high. In particular, I would suspect<br>
you are hitting offloads hard, and the current (fixed in linux 4.1)<br>
codel drop algorithm stops dropping below "maxpacket", which was meant<br>
in the ns2 code to be a MTU, but in the linux code ended up being a<br>
TSO sized (64k!) packet.<br>
<br>
tc -s qdisc show # will show the maxpacket.<br>
<span class=""><br>
> Tell me if there is more information I can provide or tests to run.<br>
<br>
</span>0) I tend to have a script for everything... I can give you an example<br>
script in another email.<br>
<br>
1) tc -s qdisc show dev whatever # will show actuall drop/mark statistics<br>
<br>
2) Please run your flent tests with -x --disable-log<br>
<br>
-x collects more metadata, --disable-log disables the automatic log<br>
scaling which is so hard to discern on these plots.<br>
<br>
Use -t "title" to differentiate between variables under test.<br>
<br>
3) I also tend to use flent's --remote-metadata=root@your_openwrtbox<br>
to get the stats on that box into the metadata. You have to add your<br>
local .ssh/id_rsa.pub key to<br>
your_openwrtbox:/etc/dropbear/authorized_keys file to do this.<br>
<br>
4) With all that in hand, sticking up a tarball of the results makes<br>
for easy plotting of various other graphs, and using the flent-gui,<br>
you can combine results from each run easily, also.<br>
<br>
5) try disabling offloads on all interfaces on the router (or running cake)<br>
<br>
My usual suite of tests is rrul, rrul_be, tcp_1up, tcp_1down, and tcp_2up_delay.<br>
<br>
and rtt_fair (if you have more than one target server available)...<br>
all without the iperf stuff....<br>
<br>
Your mss changing idea is interesting, and I would like to do a mod to<br>
tcp_2up_delay to sort of match your iperf combination + flent to<br>
totally capture all data.<br>
<br>
6) I am pretty interested as to what happens *without* sqm at the max<br>
forwarding rate with fq_codel engaged on all these tests.<br>
<span class="im HOEnZb"><br>
><br>
> --<br>
> Mikael Abrahamsson email: <a href="mailto:swmike@swm.pp.se">swmike@swm.pp.se</a><br>
><br>
</span><span class="im HOEnZb">> _______________________________________________<br>
> Cerowrt-devel mailing list<br>
> <a href="mailto:Cerowrt-devel@lists.bufferbloat.net">Cerowrt-devel@lists.bufferbloat.net</a><br>
> <a href="https://lists.bufferbloat.net/listinfo/cerowrt-devel" rel="noreferrer" target="_blank">https://lists.bufferbloat.net/listinfo/cerowrt-devel</a><br>
><br>
<br>
<br>
<br>
</span><span class="im HOEnZb">--<br>
Dave Täht<br>
worldwide bufferbloat report:<br>
<a href="http://www.dslreports.com/speedtest/results/bufferbloat" rel="noreferrer" target="_blank">http://www.dslreports.com/speedtest/results/bufferbloat</a><br>
And:<br>
What will it take to vastly improve wifi for everyone?<br>
<a href="https://plus.google.com/u/0/explore/makewififast" rel="noreferrer" target="_blank">https://plus.google.com/u/0/explore/makewififast</a><br>
</span><div class="HOEnZb"><div class="h5">_______________________________________________<br>
Cerowrt-devel mailing list<br>
<a href="mailto:Cerowrt-devel@lists.bufferbloat.net">Cerowrt-devel@lists.bufferbloat.net</a><br>
<a href="https://lists.bufferbloat.net/listinfo/cerowrt-devel" rel="noreferrer" target="_blank">https://lists.bufferbloat.net/listinfo/cerowrt-devel</a><br>
</div></div></blockquote></div><br></div>