[Cerowrt-devel] DSLReports Speed Test has latency measurement built-in

[Cerowrt-devel] DSLReports Speed Test has latency measurement built-in

[Rich Brown]

Folks,

I am delighted to pass along the news that Justin has added latency measurements into the Speed Test at DSLReports.com.

Go to: https://www.dslreports.com/speedtest and click the button for your Internet link. This controls the number of simultaneous connections that get established between your browser and the speedtest server. After you run the test, click the green "Results + Share" button to see detailed info. For the moment, you need to be logged in to see the latency results. There's a "register" link on each page.

The speed test measures latency using websocket pings: Justin says that a zero-latency link can give 1000 Hz - faster than a full HTTP ping. I just ran a test and got 48 msec latency from DSLReports, while ping gstatic.com gave 38-40 msec, so they're pretty fast. 

You can leave feedback on this page - http://www.dslreports.com/forum/r29910594-FYI-for-general-feedback-on-the-new-speedtest - or wait 'til Justin creates a new Bufferbloat topic on the forums.

Enjoy!

Rich

Re: [Cerowrt-devel] [Bloat] DSLReports Speed Test has latency measurement built-in

[Dave Taht]

What I see here is the same old latency, upload, download series, not
latency and bandwidth at the same time.

http://www.dslreports.com/speedtest/319616

On Sat, Apr 18, 2015 at 5:57 PM, Rich Brown <richb.hanover@gmail.com> wrote:
> Folks,
>
> I am delighted to pass along the news that Justin has added latency measurements into the Speed Test at DSLReports.com.
>
> Go to: https://www.dslreports.com/speedtest and click the button for your Internet link. This controls the number of simultaneous connections that get established between your browser and the speedtest server. After you run the test, click the green "Results + Share" button to see detailed info. For the moment, you need to be logged in to see the latency results. There's a "register" link on each page.
>
> The speed test measures latency using websocket pings: Justin says that a zero-latency link can give 1000 Hz - faster than a full HTTP ping. I just ran a test and got 48 msec latency from DSLReports, while ping gstatic.com gave 38-40 msec, so they're pretty fast.
>
> You can leave feedback on this page - http://www.dslreports.com/forum/r29910594-FYI-for-general-feedback-on-the-new-speedtest - or wait 'til Justin creates a new Bufferbloat topic on the forums.
>
> Enjoy!
>
> Rich
> _______________________________________________
> Bloat mailing list
> Bloat@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/bloat



-- 
Dave Täht
Open Networking needs **Open Source Hardware**

https://plus.google.com/u/0/+EricRaymond/posts/JqxCe2pFr67

Re: [Cerowrt-devel] DSLReports Speed Test has latency measurement built-in

[leetminiwheat]

On Sat, Apr 18, 2015 at 8:57 PM, Rich Brown <richb.hanover@gmail.com> wrote:
>
> Folks,
>
> I am delighted to pass along the news that Justin has added latency measurements into the Speed Test at DSLReports.com.
>
> Go to: https://www.dslreports.com/speedtest and click the button for your Internet link. This controls the number of simultaneous connections that get established between your browser and the speedtest server. After you run the test, click the green "Results + Share" button to see detailed info. For the moment, you need to be logged in to see the latency results. There's a "register" link on each page.
>
> The speed test measures latency using websocket pings: Justin says that a zero-latency link can give 1000 Hz - faster than a full HTTP ping. I just ran a test and got 48 msec latency from DSLReports, while ping gstatic.com gave 38-40 msec, so they're pretty fast.
>
> You can leave feedback on this page - http://www.dslreports.com/forum/r29910594-FYI-for-general-feedback-on-the-new-speedtest - or wait 'til Justin creates a new Bufferbloat topic on the forums.
Interesting! Thanks for sharing.

Tested on my win box w/ tweaked TCP settings and tweaked intel NIC,
and also on my linux laptop. strange the results differed quite a bit.
My win box had a lot of Re-Xmit and jitter while my linux box had no
Re-Xmit and extremely low jitter. strange, maybe lowering intel NIC
buffers from Rx:256/Tx:512 to Rx:128/Tx:256 had a negative effect.
This machine is a monster too, i7 4790k, 32GB 2400MHz DDR3, GTX TITAN
X, meanwhile my laptop is an old gen1 core i5 with crappy Realtek NIC.

tests on verizon FiOS 32/25 limited to 90%, Cero 3.10.50-1

Windows 8.1/Chrome/CTCP/window scaling enabled
-------------------------------------------------------------------------------------
0.1s start test for Fiber
0.41s 10 locations to query
9.46s latency New York, NY, USA is avg 7ms
9.46s latency Atlanta, GA is avg 27ms
9.46s latency Chicago, IL, USA is avg 29ms
9.46s latency Council Bluffs, Iowa, USA is avg 32ms
9.46s latency Charlottesville, Virgina, USA is avg 43ms
9.46s latency Dallas, TX, USA is avg 48ms
9.46s latency Denver, Colorado, USA is avg 53ms
9.46s latency Seattle, WA, USA is avg 70ms
9.46s latency Los Angeles, CA, USA is avg 70ms
9.46s latency San Francisco, CA, USA is avg 79ms
11.47s Starting download phase
19.65s  stream0 mbits=0.94 (5% of 64k buffer) Atlanta, GA 1sec=1 9%
19.65s  stream1 mbits=0.94 (5% of 64k buffer) Atlanta, GA 1sec=0.9 0%
19.65s  stream2 mbits=0.96 (1% of 64k buffer) New York, NY, USA 1sec=0.85 0%
19.65s  stream3 mbits=0.96 (5% of 64k buffer) Atlanta, GA 1sec=0.97 0%
19.65s  stream4 mbits=0.91 (5% of 64k buffer) Chicago, IL, USA 1sec=1 0%
19.65s  stream5 mbits=0.95 (1% of 64k buffer) New York, NY, USA 1sec=0.88 0%
19.65s  stream6 mbits=0.93 (5% of 64k buffer) Chicago, IL, USA 1sec=0.91 0%
19.65s  stream7 mbits=0.98 (1% of 64k buffer) New York, NY, USA 1sec=1.04 0%
19.65s  stream8 mbits=0.92 (5% of 64k buffer) Chicago, IL, USA 1sec=0.88 0%
19.65s  stream9 mbits=0.89 (5% of 64k buffer) Council Bluffs, Iowa,
USA 1sec=0.81 0%
19.65s  stream10 mbits=0.92 (5% of 64k buffer) Atlanta, GA 1sec=0.93 0%
19.65s  stream11 mbits=0.89 (5% of 64k buffer) Council Bluffs, Iowa,
USA 1sec=0.9 0%
19.65s  stream12 mbits=0.9 (5% of 64k buffer) Chicago, IL, USA 1sec=0.93 0%
19.65s  stream13 mbits=0.97 (1% of 64k buffer) New York, NY, USA 1sec=0.99 0%
19.65s  stream14 mbits=0.86 (5% of 64k buffer) Chicago, IL, USA 1sec=0.86 0%
19.65s  stream15 mbits=0.58 (1% of 64k buffer) New York, NY, USA 1sec=0.35 0%
19.65s  stream16 mbits=0.41 (2% of 64k buffer) Chicago, IL, USA 1sec=0.44 0%
19.65s  stream17 mbits=0.87 (4% of 64k buffer) Atlanta, GA 1sec=0.88 0%
19.65s  stream18 mbits=0.88 (5% of 64k buffer) Council Bluffs, Iowa,
USA 1sec=0.83 0%
19.65s  stream19 mbits=0.92 (5% of 64k buffer) Chicago, IL, USA 1sec=0.86 0%
19.65s  stream20 mbits=0.89 (5% of 64k buffer) Council Bluffs, Iowa,
USA 1sec=0.95 0%
19.65s  stream21 mbits=0.86 (5% of 64k buffer) Chicago, IL, USA 1sec=0.93 0%
19.65s  stream22 mbits=0.54 (1% of 64k buffer) New York, NY, USA 1sec=0.58 0%
19.65s  stream23 mbits=0.89 (5% of 64k buffer) Council Bluffs, Iowa,
USA 1sec=0.77 0%
19.65s  stream24 mbits=0.4 (2% of 64k buffer) Atlanta, GA 1sec=0.49 0%
19.65s  stream25 mbits=0.92 (6% of 64k buffer) Council Bluffs, Iowa,
USA 1sec=0.59 0%
19.65s  stream26 mbits=0.86 (5% of 64k buffer) Council Bluffs, Iowa,
USA 1sec=0.92 0%
19.65s  stream27 mbits=0.95 (1% of 64k buffer) New York, NY, USA 1sec=1.04 0%
19.65s  stream28 mbits=0.96 (1% of 64k buffer) New York, NY, USA 1sec=0.9 0%
19.65s  stream29 mbits=0.88 (5% of 64k buffer) Council Bluffs, Iowa,
USA 1sec=0.99 0%
19.65s  stream30 mbits=0.94 (1% of 64k buffer) New York, NY, USA 1sec=0.85 0%
19.65s  stream31 mbits=0.91 (5% of 64k buffer) Atlanta, GA 1sec=0.88 0%
19.65s download total=27.6
22.95s Starting upload phase
31.06s  stream0 (Atlanta, GA) mbits=3.68
31.06s  stream1 (Atlanta, GA) mbits=3.82
31.06s  stream2 (New York, NY, USA) mbits=3.94
31.06s  stream3 (Atlanta, GA) mbits=3.68
31.06s  stream4 (Chicago, IL, USA) mbits=3.66
31.06s  stream5 (New York, NY, USA) mbits=3.82
31.06s upload total=22.6
44.85s END TEST
44.85s 10hz drop counts f=1 ms=201
48.99s Total megabytes used: 96.2 (down:36.5 up:59.8)

Flow Stats:
4ms-17ms jitter
4%-15% Re-Xmit
3-10 Cwnd on every test.

-------------------------------------------------------------------------------------

Ubuntu-MATE/Chromium/fq_codel/window scaling enabled
0.11s start test for Fiber
0.26s 10 locations to query
9.31s latency New York, NY, USA is avg 7ms
9.31s latency Atlanta, GA is avg 27ms
9.31s latency Council Bluffs, Iowa, USA is avg 33ms
9.31s latency Chicago, IL, USA is avg 34ms
9.31s latency Charlottesville, Virgina, USA is avg 41ms
9.31s latency Dallas, TX, USA is avg 48ms
9.31s latency Denver, Colorado, USA is avg 55ms
9.31s latency Los Angeles, CA, USA is avg 71ms
9.31s latency Seattle, WA, USA is avg 71ms
9.31s latency San Francisco, CA, USA is avg 78ms
11.31s Starting download phase
19.82s  stream0 mbits=1.53 (2% of 64k buffer) New York, NY, USA 1sec=1.56 16%
19.83s  stream1 mbits=1.52 (2% of 64k buffer) New York, NY, USA 1sec=1.56 1%
19.83s  stream2 mbits=1.5 (8% of 64k buffer) Atlanta, GA 1sec=1.6 1%
19.83s  stream3 mbits=1.48 (8% of 64k buffer) Atlanta, GA 1sec=1.57 1%
19.83s  stream4 mbits=1.48 (8% of 64k buffer) Atlanta, GA 1sec=1.57 1%
19.83s  stream5 mbits=1.49 (8% of 64k buffer) Atlanta, GA 1sec=1.57 1%
19.83s  stream6 mbits=1.48 (8% of 64k buffer) Atlanta, GA 1sec=1.55 1%
19.83s  stream7 mbits=1.48 (8% of 64k buffer) Atlanta, GA 1sec=1.57 1%
19.83s  stream8 mbits=1.53 (2% of 64k buffer) New York, NY, USA 1sec=1.56 1%
19.83s  stream9 mbits=1.51 (2% of 64k buffer) New York, NY, USA 1sec=1.56 1%
19.83s  stream10 mbits=1.53 (2% of 64k buffer) New York, NY, USA 1sec=1.59 1%
19.83s  stream11 mbits=1.51 (2% of 64k buffer) New York, NY, USA 1sec=1.56 1%
19.83s  stream12 mbits=1.47 (8% of 64k buffer) Atlanta, GA 1sec=1.57 1%
19.83s  stream13 mbits=1.51 (2% of 64k buffer) New York, NY, USA 1sec=1.59 1%
19.83s  stream14 mbits=1.47 (8% of 64k buffer) Atlanta, GA 1sec=1.57 1%
19.83s  stream15 mbits=1.51 (2% of 64k buffer) New York, NY, USA 1sec=1.56 1%
19.83s  stream16 mbits=1.49 (8% of 64k buffer) Atlanta, GA 1sec=1.57 1%
19.83s  stream17 mbits=1.51 (2% of 64k buffer) New York, NY, USA 1sec=1.56 1%
19.83s download total=27
23.12s Starting upload phase
31.35s  stream0 (New York, NY, USA) mbits=3.64
31.35s  stream1 (New York, NY, USA) mbits=3.64
31.35s  stream2 (Atlanta, GA) mbits=3.64
31.35s  stream3 (Atlanta, GA) mbits=3.64
31.35s  stream4 (Atlanta, GA) mbits=3.64
31.35s  stream5 (Atlanta, GA) mbits=3.64
31.35s upload total=21.8
45.58s END TEST
45.58s 10hz drop counts f=2 ms=471
49.71s Total megabytes used: 97.1 (down:36.9 up:60.2)

Flow Stats:
0.3ms-5.9ms jitter
0% Re-xmit
2-6 Cwnd on every test, mostly 2
-------------------------------------------------------------------------------------

Re: [Cerowrt-devel] [Bloat] DSLReports Speed Test has latency measurement built-in

[Dave Taht]

This test was taken on linux, about 20 feet and one room away from the
access point:

http://www.dslreports.com/speedtest/320328

This was taken on the same box, about 10 feet and one room from the
access point.

http://www.dslreports.com/speedtest/320340

In all cases, the uplink is a comcast box configured for 55Mbit down,
5Mbit up and just to make it weird this is a two router configuration,
where the nearest hop is over a powerline box (TP600) before hitting
the net.

I *like* that the test does not let you switch browser tabs (something
I do instinctively when something takes longer than 3 seconds.)




On Sat, Apr 18, 2015 at 5:57 PM, Rich Brown <richb.hanover@gmail.com> wrote:
> Folks,
>
> I am delighted to pass along the news that Justin has added latency measurements into the Speed Test at DSLReports.com.
>
> Go to: https://www.dslreports.com/speedtest and click the button for your Internet link. This controls the number of simultaneous connections that get established between your browser and the speedtest server. After you run the test, click the green "Results + Share" button to see detailed info. For the moment, you need to be logged in to see the latency results. There's a "register" link on each page.
>
> The speed test measures latency using websocket pings: Justin says that a zero-latency link can give 1000 Hz - faster than a full HTTP ping. I just ran a test and got 48 msec latency from DSLReports, while ping gstatic.com gave 38-40 msec, so they're pretty fast.
>
> You can leave feedback on this page - http://www.dslreports.com/forum/r29910594-FYI-for-general-feedback-on-the-new-speedtest - or wait 'til Justin creates a new Bufferbloat topic on the forums.
>
> Enjoy!
>
> Rich
> _______________________________________________
> Bloat mailing list
> Bloat@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/bloat



-- 
Dave Täht
Open Networking needs **Open Source Hardware**

https://plus.google.com/u/0/+EricRaymond/posts/JqxCe2pFr67

Re: [Cerowrt-devel] [Bloat] DSLReports Speed Test has latency measurement built-in

[Dave Taht]

This was a test taken *during* a 2 minute rrul_be test.

http://www.dslreports.com/speedtest/320377

Flient (formerly netperf-wrapper) data here:
http://snapon.lab.bufferbloat.net/~d/lorna-wifi.tgz

Puzzle over this!
http://snapon.lab.bufferbloat.net/~d/lorna-wifi/reconcile_this.png and
the rawer data.... in comparison to this and other of these new
speedtest reports.

There are a couple other tests of the same link in the  same
configuration (laptop on lap 10 feet from the access point through a
wall) [1] in the same dir testing upload and download (without
simultaneously running the new dslreport tests)

CDF plots are nice. So are mountain plots.

http://snapon.lab.bufferbloat.net/~d/lorna-wifi/wifi_download.png






[1] I was trying for comfort^H^H^H^H^^H^H^H^Hrealism

On Sun, Apr 19, 2015 at 1:29 AM, Dave Taht <dave.taht@gmail.com> wrote:
> This test was taken on linux, about 20 feet and one room away from the
> access point:
>
> http://www.dslreports.com/speedtest/320328
>
> This was taken on the same box, about 10 feet and one room from the
> access point.
>
> http://www.dslreports.com/speedtest/320340
>
> In all cases, the uplink is a comcast box configured for 55Mbit down,
> 5Mbit up and just to make it weird this is a two router configuration,
> where the nearest hop is over a powerline box (TP600) before hitting
> the net.
>
> I *like* that the test does not let you switch browser tabs (something
> I do instinctively when something takes longer than 3 seconds.)
>
>
>
>
> On Sat, Apr 18, 2015 at 5:57 PM, Rich Brown <richb.hanover@gmail.com> wrote:
>> Folks,
>>
>> I am delighted to pass along the news that Justin has added latency measurements into the Speed Test at DSLReports.com.
>>
>> Go to: https://www.dslreports.com/speedtest and click the button for your Internet link. This controls the number of simultaneous connections that get established between your browser and the speedtest server. After you run the test, click the green "Results + Share" button to see detailed info. For the moment, you need to be logged in to see the latency results. There's a "register" link on each page.
>>
>> The speed test measures latency using websocket pings: Justin says that a zero-latency link can give 1000 Hz - faster than a full HTTP ping. I just ran a test and got 48 msec latency from DSLReports, while ping gstatic.com gave 38-40 msec, so they're pretty fast.
>>
>> You can leave feedback on this page - http://www.dslreports.com/forum/r29910594-FYI-for-general-feedback-on-the-new-speedtest - or wait 'til Justin creates a new Bufferbloat topic on the forums.
>>
>> Enjoy!
>>
>> Rich
>> _______________________________________________
>> Bloat mailing list
>> Bloat@lists.bufferbloat.net
>> https://lists.bufferbloat.net/listinfo/bloat
>
>
>
> --
> Dave Täht
> Open Networking needs **Open Source Hardware**
>
> https://plus.google.com/u/0/+EricRaymond/posts/JqxCe2pFr67



-- 
Dave Täht
Open Networking needs **Open Source Hardware**

https://plus.google.com/u/0/+EricRaymond/posts/JqxCe2pFr67

Re: [Cerowrt-devel] [Bloat] DSLReports Speed Test has latency measurement built-in

[MUSCARIELLO Luca IMT/OLN]

Nice tool.

Here is a test from France.
http://www.dslreports.com/speedtest/320422

There are not so many servers in Europe I guess.



On 04/19/2015 10:29 AM, Dave Taht wrote:
> This test was taken on linux, about 20 feet and one room away from the
> access point:
>
> http://www.dslreports.com/speedtest/320328
>
> This was taken on the same box, about 10 feet and one room from the
> access point.
>
> http://www.dslreports.com/speedtest/320340
>
> In all cases, the uplink is a comcast box configured for 55Mbit down,
> 5Mbit up and just to make it weird this is a two router configuration,
> where the nearest hop is over a powerline box (TP600) before hitting
> the net.
>
> I *like* that the test does not let you switch browser tabs (something
> I do instinctively when something takes longer than 3 seconds.)
>
>
>
>
> On Sat, Apr 18, 2015 at 5:57 PM, Rich Brown <richb.hanover@gmail.com> wrote:
>> Folks,
>>
>> I am delighted to pass along the news that Justin has added latency measurements into the Speed Test at DSLReports.com.
>>
>> Go to: https://www.dslreports.com/speedtest and click the button for your Internet link. This controls the number of simultaneous connections that get established between your browser and the speedtest server. After you run the test, click the green "Results + Share" button to see detailed info. For the moment, you need to be logged in to see the latency results. There's a "register" link on each page.
>>
>> The speed test measures latency using websocket pings: Justin says that a zero-latency link can give 1000 Hz - faster than a full HTTP ping. I just ran a test and got 48 msec latency from DSLReports, while ping gstatic.com gave 38-40 msec, so they're pretty fast.
>>
>> You can leave feedback on this page - http://www.dslreports.com/forum/r29910594-FYI-for-general-feedback-on-the-new-speedtest - or wait 'til Justin creates a new Bufferbloat topic on the forums.
>>
>> Enjoy!
>>
>> Rich
>> _______________________________________________
>> Bloat mailing list
>> Bloat@lists.bufferbloat.net
>> https://lists.bufferbloat.net/listinfo/bloat
>
>

Re: [Cerowrt-devel] [Bloat] DSLReports Speed Test has latency measurement built-in

[Colin Dearborn]

If you go to the results page, you get a "ping time during test" graph.


-----Original Message-----
From: bloat-bounces@lists.bufferbloat.net [mailto:bloat-bounces@lists.bufferbloat.net] On Behalf Of Dave Taht
Sent: Saturday, April 18, 2015 10:02 PM
To: Rich Brown
Cc: cerowrt-devel; bloat
Subject: Re: [Bloat] DSLReports Speed Test has latency measurement built-in

What I see here is the same old latency, upload, download series, not latency and bandwidth at the same time.

http://www.dslreports.com/speedtest/319616

On Sat, Apr 18, 2015 at 5:57 PM, Rich Brown <richb.hanover@gmail.com> wrote:
> Folks,
>
> I am delighted to pass along the news that Justin has added latency measurements into the Speed Test at DSLReports.com.
>
> Go to: https://www.dslreports.com/speedtest and click the button for your Internet link. This controls the number of simultaneous connections that get established between your browser and the speedtest server. After you run the test, click the green "Results + Share" button to see detailed info. For the moment, you need to be logged in to see the latency results. There's a "register" link on each page.
>
> The speed test measures latency using websocket pings: Justin says that a zero-latency link can give 1000 Hz - faster than a full HTTP ping. I just ran a test and got 48 msec latency from DSLReports, while ping gstatic.com gave 38-40 msec, so they're pretty fast.
>
> You can leave feedback on this page - http://www.dslreports.com/forum/r29910594-FYI-for-general-feedback-on-the-new-speedtest - or wait 'til Justin creates a new Bufferbloat topic on the forums.
>
> Enjoy!
>
> Rich
> _______________________________________________
> Bloat mailing list
> Bloat@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/bloat



--
Dave Täht
Open Networking needs **Open Source Hardware**

https://plus.google.com/u/0/+EricRaymond/posts/JqxCe2pFr67
_______________________________________________
Bloat mailing list
Bloat@lists.bufferbloat.net
https://lists.bufferbloat.net/listinfo/bloat