<div dir="ltr"><div>Regarding this part:</div><div><br></div>> <span style="font-size:13px">The baseline latency is because the bloat measurement uses a single websocket ping server in America. Justin said in the forums it didn't seem worth the effort to set up more of them. Seems worth an faq item though :(.</span><div><br></div><div>Below huge speeds, upload testing is done with web socket now,</div><div>so there is a websocket address on every server now anyway.</div><div><br></div><div>The baseline pinging to <a href="http://dslreports.com">dslreports.com</a> doesn't seem broken</div><div>but if necessary it can be changed to baseline pinging to the</div><div>nearest server, wherever that is.</div><div><div class="gmail_extra"><br><div class="gmail_quote">On Sat, May 23, 2015 at 12:05 AM, Alan Jenkins <span dir="ltr"><<a href="mailto:alan.christopher.jenkins@gmail.com" target="_blank">alan.christopher.jenkins@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="HOEnZb"><div class="h5">On 20/05/15 13:47, Kevin Darbyshire-Bryant wrote:<br>
> On 19/05/2015 23:37, Dave Taht wrote:<br>
>><br>
>> 0) dslreports has a hires bufferbloat option now in their settings.<br>
>> It reveals much detail that I like very much. It may not work well<br>
>> on some browsers. Give it a shot, please.<br>
> Tried it - fun! Here are some of the results of that fun, I've no<br>
> idea if this will help anyone.<br>
>><br>
>> 1) I like that the graphic .png reports now a ping range, but I<br>
>> think that is baseline latencies. but I think it would be clearer<br>
>> if it showed the up and the down, under load, 98th percentile,<br>
>> also.<br>
>><br>
>> an unshaped, unmodified cable modem result in all it's horrible<br>
>> glory:<br>
>><br>
>> <a href="http://www.dslreports.com/speedtest/506793" target="_blank">http://www.dslreports.com/speedtest/506793</a><br>
><br>
> Ok so this is my HG612 unshaped VDSL2 modem sat behind Archer C7.<br>
> The VDSL2 link is 40Mbit down, 10Mbit up as reported as its capped<br>
> 'sync' rate.<br>
><br>
> <a href="http://www.dslreports.com/speedtest/513588" target="_blank">http://www.dslreports.com/speedtest/513588</a><br>
><br>
> Up is horrendous and I think you can see classic TCP sawtoothing in<br>
> the delay as well. Peak just as 'up test' goes idle is strange<br>
> though.<br>
><br>
>> 2) the "cable" test (which keeps changing the number of flows -<br>
>> these are all 16/6 flows) thoroughly breaks the sqm system's<br>
>> inbound rate shaper, using cake or fq_codel (cake here), with my<br>
>> rate set 12% below the delivered rate (100mbit vs 112Mbit).<br>
><br>
> Behaviour here is different. In theory same test 16/6 flows, down is<br>
> capped 38Mbit (5%), up at 9.7Mbit(3%) These are all 'cake'<br>
><br>
> <a href="http://www.dslreports.com/speedtest/513627" target="_blank">http://www.dslreports.com/speedtest/513627</a><br>
><br>
> There's a slight double bump at the beginning of the down latency<br>
> test, but otherwise my browsing/upload/download experience is vastly<br>
> improved.<br>
><br>
> Tuning the down to 37Mbit helps that bump a little maybe:<br>
> <a href="http://www.dslreports.com/speedtest/513652" target="_blank">http://www.dslreports.com/speedtest/513652</a><br>
><br>
> It gets really interesting if I increase the down limit to 39Mbit:<br>
> <a href="http://www.dslreports.com/speedtest/513782" target="_blank">http://www.dslreports.com/speedtest/513782</a> where it starts to look a<br>
> bit like your test results.<br>
><br>
> Increasing the up limit showed an interesting step change in upload<br>
> delay, this is the up cap set to 9.8Mbit<br>
> <a href="http://www.dslreports.com/speedtest/513863" target="_blank">http://www.dslreports.com/speedtest/513863</a> (I had a really good<br>
> example of the delay increasing in a linear fashion as the buffer<br>
> just couldn't drain fast enough...if I find that test result again<br>
> I'll post it)<br>
><br>
> Tuning it back down, by even as little as 50kbits would remove that<br>
> step, I settled on 9.7Mbit for safety.<br>
><br>
> The elephant in my personal room is the high latency baseline<br>
> measurement. None of the ping response time test sites I've checked<br>
> give me anywhere near a baseline ping rtt of 100ms. Even dslreports<br>
> say "London UK is ~10ms, Google Europe is ~17ms, Dublin, Ireland, EU<br>
> is ~20ms, Frankfurt, DE, EU is ~27ms" So I clearly don't<br>
> understand some thing(s) about this test.<br>
><br>
> Anyway, that's been an interesting 2 hours of playing!<br>
><br>
> Kevin<br>
<br></div></div>
Good fun :).<br>
<br>
The baseline latency is because the bloat measurement uses a single websocket ping server in America. Justin said in the forums it didn't seem worth the effort to set up more of them. Seems worth an faq item though :(.<span class="HOEnZb"><font color="#888888"><br>
<br>
Alan</font></span><div class="HOEnZb"><div class="h5"><br>
<br>
_______________________________________________<br>
Bloat mailing list<br>
<a href="mailto:Bloat@lists.bufferbloat.net" target="_blank">Bloat@lists.bufferbloat.net</a><br>
<a href="https://lists.bufferbloat.net/listinfo/bloat" target="_blank">https://lists.bufferbloat.net/listinfo/bloat</a><br>
</div></div></blockquote></div><br></div></div></div>