MLAB Offers 'Internet Health Test' For Neutrality Violations

Thanks, already there.

Make AT&T, Verizon, Comcast, Suddenlink, CenturyLink, Frontier, Fairpoint, Windstream, Cox, and Tim Warner nervous. They should be.

You forgot Mediacom but that's OK I really hope someone comes in and buys them out. With 85% of Iowa something needs to change be Cox, Comcast, Time Warner? We are the forgotten state of Broadband for sure

What about charter?

Nervous about what? I'm still curious what is really going to be so magically different after 13 June.

I have a 10Gig pipe and this is the results. »bit.ly/1AdKuG9 How insulting to get this response.
I tried it a few times without any better results. Data centre fail?
I had to add one from our fiber connection, It shows better. All I have is a gig net card on my computer.

How well this works depends on how many mlab nodes there are. They don't say. And they only have 3 servers per node to handle tests. My results were all over the place. But my dslreports speed tests and www.speedtest.net tests are consistently excellent.

I will have to be cautious whether mlab results are actually worth anything in measuring net neutrality compliance by ISPs.

At least it is consistent, this is what I get on 1 gig: »bit.ly/1ee5LW8

Nice idea but test doesn't work. Just stays on preparing. Never goes to step 2.

Had that problem once. Fixed when I closed the window and tried a new test.

I get a blank page on Fx 38.

I hate any group that is so backward they force you to use IE. (My airlines has a new website that works only with IE). Anyhow, I tried IE 10 on Win 8 Pro and that was a worse experience than on Fx. I had to allow a bunch of cookies and then I got a page that said "thanks for doing the test". Then insultingly it presented me two big buttons to share on Twitter and Facebook but NO test ran and NO results were shown to me but I guess if I wanted to share that others could see the results I couldn't see.

I closed the tab and then closed IE and, lo and behold, there was a cookie prompt UNDERNEATH the page AND underneath it was a POPUP WINDOW that was blank but I suppose if that cookie had appeared on top where I would have approved it (after checking to make sure it wasn't a nasty cookie) then I guess a POPUP would have appeared and actually run the test. What a JOKE and PATHETIC for mlabs as they generally know something about how to setup tools and tests. Maybe its all IE's fault but I only tried IE after Fx gave me a blank page.

Firefox wouldn't start the test past 'preparing', so I switched to Chrome and it worked.

Decent little test.

ran the test 10 times (6 IE, 4 chrome) ALL had 3 sub 8ms that beat 110/11.5 (I'm on 105/10)
and ALL had one high latency 54-100ms loser (as low as 23/9)
The consistently bad one was GTT (IBM) whose network maps show at least 6 wholly owned routes to Seattle from US locations. and 3 POP's in seattle.. they are either really short on bandwidth or not doing something very well.

Have a 15/2 connection, although it often rates a tad bit higher.

»bit.ly/1Ak3Ysx

AT&T DSL. Am supposed to get 6 megs. real speed 4.32 high / 4.1 low.

This isn't a surprise, but Sonic maxed my connection in all four tests. It's nice to have an ISP that cares about this stuff instead of trying to profit off of it.

For steps 1-3, I had a download speed of over 20mbits. With Step 4 I had 1.09 Mbits. Apprently its a company called GTT thats getting 1.09 mbits.

The test seems to reflect the slowness I observe when accessing my colo which resides Cogent transit.

How are ISP's and interconnection points prevented from giving preferential treatment to the test traffic to create an all is well façade?

»bit.ly/1diFt52

This test is a complete joke. I ran it from my house (1Gig) and from servers in two Data Centers with 10Gb connectivity with a lot of direct in and indirect peers in the same router. I couldn't get above 15Mb at best not to mention the Listed Transit providers were incorrect and misleading.

The loop isn't just the ISP's last mile, that is easy to police but was never the bottleneck. Once data is dumped in the loop beyond the last mile many and various provider many or may not be involved. The FCC admitted they are totally clueless if they try to grab and hold the cloud responsible.

The best thing that ever happened to buck passers was deregulation.

Edit: I finally got the test to work with IE. I have a problem, possible, in two connections. Great now they are in trouble.

»bit.ly/1Sa4wab

I finally got this to work sort of on IE 10 on Windows 8 Pro but IE says the test is designed for OLDER BROWSERS ONLY and only worked when I used the compatibility button for IE.

I got download only results but the test did not return any error messages and IE said AGAIN that the test is designed for OLD browsers. I could see it testing the upload speed but the results show only the download speed.

»bit.ly/1Pw3MNI

Passed all but one with acceptable results on my 75/75 FiOS connection in NYC

Too bad it doesn't show/say the faulty point/connection/provider.

Google Chrome did the whole test

Here are my results using Chrome:
»bit.ly/1Pw3RRy

OK I get a better connection late at night. »bit.ly/1cOpoTS

This proves the more people that use the Interweb the slower it goes; or the "lab's" servers may be overloaded. You all should donate money like they ask so they can buy more servers.

Firefox even works at night.
»bit.ly/1AkOAvV

The FCC must demand that this "lab" install enough servers to handle the demand during prime time even if they remain idle the rest of the day; that way we eliminate the "lab" as the problem.

i don't feel "safe" with mlab or even dslr's new test, because i have to disable a lot of privacy protection chrome extensions to get either test to work right.

i'll stick with using testmy.net - if you don't mind wher eyou can judge "consistancy" by running an auto repeat test or test with a larger data set.

Here are my results:

»bit.ly/1HgPLKa

Playing with K-Meleon 74 and can't get past preparing test either!