<div dir="ltr"><div class="gmail_default" style="font-size:small"><br></div><div class="gmail_extra"><br><br><div class="gmail_quote">On Thu, Mar 21, 2013 at 2:14 PM, Oliver Hohlfeld <span dir="ltr"><<a href="mailto:oliver@net.t-labs.tu-berlin.de" target="_blank">oliver@net.t-labs.tu-berlin.de</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div class="im">> I believe you are actually measuring the *fraction of the time* your<br>
> measurements show bad latency<br>
<br>
</div>Yes.<br>
<div class="im"><br>
> The best data I've seen on how widespread the problem is is the ICSI<br>
> Netalyzr scatter plots results<br>
<br>
</div>One has to be extremely careful on what to conclude from this data.<br>
The Netalyzr data shows that bloated buffers _exist_, not that they<br>
are _used_ in practice. Or as Mark has put it: "[it] shows that large<br>
delays due to buffering can happen, not that they do happen."<br>
Empirical evidence cited in my previous mail suggests that buffer<br>
bloat does not happen often.<br></blockquote><div><br></div><div style="font-size:small" class="gmail_default">Heh. </div><div style="font-size:small" class="gmail_default"><br></div><div style="font-size:small" class="gmail_default">
6% of the time isn't "often"? The measurements is much more a measurement of how much of the time your competing with someone else (or yourself in background) for the bottleneck than anything else.</div><div style="font-size:small" class="gmail_default">
<br></div><div style="font-size:small" class="gmail_default">Start up any long lived TCP connection (using a modern TCP; that leaves out Windows XP).</div><div style="font-size:small" class="gmail_default"><br></div><div style="font-size:small" class="gmail_default">
Stand back a few seconds.</div><div style="font-size:small" class="gmail_default"><br></div><div style="font-size:small" class="gmail_default">Measure latency.</div><div style="font-size:small" class="gmail_default"><br></div>
<div style="font-size:small" class="gmail_default">Think uploading/downloading files/videos, or sending email with images attached, etc.</div><div style="font-size:small" class="gmail_default"><br></div><div style="font-size:small" class="gmail_default">
<br></div><div style="font-size:small" class="gmail_default">More insidious is what visiting a web page with lots of embedded images can do.</div><div style="font-size:small" class="gmail_default"><br></div><div style="font-size:small" class="gmail_default">
The cross product of:</div><div style="font-size:small" class="gmail_default"> 1) current web browsers using many more than 2 TCP connections.</div><div style="font-size:small" class="gmail_default"> 2) web site "sharding".</div>
<div style="font-size:small" class="gmail_default"> 3) the initial window of data in those TCP connections.</div><div style="font-size:small" class="gmail_default"><br></div><div style="font-size:small" class="gmail_default">
This causes transients of what is up to hundreds of milliseconds of head of line blocking even</div><div style="font-size:small" class="gmail_default">on very high speed broadband/wifi connections. Again, I encourage you to do this measurement on your own broadband connection, by visiting any image heavy web site and running ping....</div>
<div style="font-size:small" class="gmail_default"><br></div><div style="font-size:small" class="gmail_default"></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div class="im"><br>
> I encourage everyone to run netalyzr and/or the mlabs tests for<br>
> bufferbloat on your own broadband connections, or do simple copy and<br>
> ping tests inside your own house over wifi to your local file servers....<br>
<br>
</div>While this will identify bloated buffers, it does not help in answering<br>
what fraction of Internet users actually experience the problem.<br></blockquote><div><br></div><div style="font-size:small" class="gmail_default">We've not come across any broadband equipment buffered "correctly". If thought is given at all to buffering, it's been sized to the maximum buffering the device might ever need for a single TCP flow at its maximum bandwidth. Example, plug a current DOCSIS 3 modem (capable up to 300Mbps these days) and use it at 20Mbps; unless the buffersizing amendment has been turned on (no ISP I am aware of yet does so), you'll be overbuffered by a factor of 15.</div>
<div style="font-size:small" class="gmail_default"><br></div><div style="font-size:small" class="gmail_default">And all the operating systems have problems, to one degree or another (and as they are all adjacent to wireless links these days, they also contribute to the problem.</div>
<div style="font-size:small" class="gmail_default"><br></div><div style="font-size:small" class="gmail_default">Bufferbloat just waits in hiding to get you when you try to use the network.</div><div style="font-size:small" class="gmail_default">
- Jim</div><div style="font-size:small" class="gmail_default"><br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<span class="HOEnZb"><font color="#888888"><br>
Oliver</font></span> </blockquote><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><span class="HOEnZb"><font color="#888888">
</font></span><div class="HOEnZb"><div class="h5">_______________________________________________<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>
</div></div></blockquote></div><br></div></div>