Jump to content

How do I convince my ISP TestMy.net is right?


CA3LE

Recommended Posts

Good Morning Damon,
                                     On the strength of testmy.net statistics I have mnade an enquiry into my slow throughput via my ISP.
They ahve sent an Engineer to test my line and everything checks out 100%.
I showed the engineer a speed test done with TMN (Which he hadn't heard of) and it was 1.9mbps downstream.  My line synchronisation with the exchange is 7400Kbps downstream and, if one does a speed test with the British Telecom speed test site we get 4.79Mbps!
So, my question is, how do I convince the ISP to rely on TMN stats and not the wholesellers personal site?
 
I produced some figures for the engineer, printed from the TMN history which clearly show that my transfer speeds were very good previously, i.e. between 5 and 6 Mbps normally but it's obviously easier for the British Telecom engineer to believe his own employers test site.
 
What should I say to try and convince them that your site is actually more accurate?  I'm sure there is a page of technical stuff somewhere :)
 
In the end of course, I'll lose the argument but it would be nice to present an argument in the first place :)
 
Thanks for any info,
 
Regards,
 
James

 

Hi James,

 
All the proof you need is in your stats.  Here's a couple of queries I did...
 
 
I attached screenshots.  You can clearly see that your speed has dropped recently.  You've tested to Amsterdam, Texas and Multithread.  Best you're getting is ~5Mbps.  There's a discrepancy between TMN and BT's speed test because they work differently.  I'll just say this... who do you think the telephone company's speed test is made to benefit?  TestMy.net is a third party and my servers are hosted outside of your ISPs network... just as all the sites you visit are.  The fact is, usually it's congested oversold peering at fault.  Your ISP wants you to test with routing taken out of the equation so their test servers are within their network.
 
You have the right to a third party opinion.  You've been testing at TestMy.net for a while, you have gathered a good amount of data using various testing methods.  From what I see... your connection isn't running as good as it what.  And I don't care what the DSL guys say... they want you to believe that it's beyond their control and a fault of the sites and services you use.  It's not.  They choose their providers just like you do, if their lines are oversold or congested they should be buying more peers or switching their current ones.  Instead... they count the money and pass the blame.
 
The multithread test is a great connection test, not only can it help push a connection to the max but it also has the ability to test a bunch of routes at once.  I can see in your stats that you're fully aware of that... you've tested using my CDN's and the closest servers and still can't do better than 5 Mbps.  If you're supposed to have 7400 Kbps then you're getting 68%... but that's at best.  You should get 80% at worst in my opinion.  In your case you should get no less than ~ 6 Mbps.
 
The classic single thread shows weakness.  Same test, same server, same amount of data... multithread (as your ISPs test no doubt is) can easily mask issues.  If your single thread performance is low streaming services and downloads will look much lower than the combined multithread performance.
 
:::.. Download Speed Test Result Details ..:::
Download Connection Speed:: 63871 Kbps or 63.9 Mbps 
Download Speed Test Size:: 99.1 MB or 101456 kB or 103891086 bytes
Download Binary File Transfer Speed:: 7984 kB/s or 8 MB/s
Tested At:: http://TestMy.net Version 13
Multithread Test Utilizing:: eu.testmy.net
 
:::.. Download Speed Test Result Details ..:::
Download Connection Speed:: 23388 Kbps or 23.4 Mbps 
Download Speed Test Size:: 100 MB or 102400 kB or 104857600 bytes
Download Binary File Transfer Speed:: 2923 kB/s or 2.9 MB/s
Tested At:: http://TestMy.net Version 13
TiP Measurement Summary:: Min 9.25 Mbps | Middle Avg 27.41 Mbps | Max 28.94 Mbps | 11% Variance
TiP Data Points:: 9.25 Mbps, 24.29 Mbps, 27.01 Mbps, 27.61 Mbps, 27.31 Mbps, 27.19 Mbps, 27.61 Mbps, 27.29 Mbps, 27.18 Mbps, 27.48 Mbps, 27.3 Mbps, 27.85 Mbps, 28.94 Mbps, 28.63 Mbps, 27.87 Mbps, 27.59 Mbps, 27.14 Mbps, 27.76 Mbps, 13.03 Mbps
 
I did both tests from Colorado Springs, CO USA to Amsterdam, NL.  ~5000 miles.  
 
To Dallas TX ~ 700 miles
 
:::.. Download Speed Test Result Details ..:::
Download Connection Speed:: 106509 Kbps or 106.5 Mbps 
Download Speed Test Size:: 100 MB or 102400 kB or 104857600 bytes
Download Binary File Transfer Speed:: 13314 kB/s or 13.3 MB/s
Tested At:: http://TestMy.net Version 13
TiP Measurement Summary:: Min 53.89 Mbps | Middle Avg 113.78 Mbps | Max 115.26 Mbps | 7% Variance
TiP Data Points:: 63.5 Mbps, 103.71 Mbps, 114.56 Mbps, 114.56 Mbps, 114.56 Mbps, 115.26 Mbps, 115.09 Mbps, 114.91 Mbps, 114.56 Mbps, 114.22 Mbps, 113.7 Mbps, 114.22 Mbps, 114.56 Mbps, 114.56 Mbps, 114.39 Mbps, 113.53 Mbps, 113.53 Mbps, 114.39 Mbps, 53.89 Mbps
 
Multithread Google CDN
 
:::.. Download Speed Test Result Details ..:::
Download Connection Speed:: 112788 Kbps or 112.8 Mbps 
Download Speed Test Size:: 200.9 MB or 205692 kB or 210628710 bytes
Download Binary File Transfer Speed:: 14099 kB/s or 14.1 MB/s
Tested At:: http://TestMy.net Version 13
Multithread Test Utilizing:: google.testmy.net
 
1/3 the speed in single thread at that distance... can you see how if you only saw the multithread result it might be misleading.  ISPs want you to only see the combined efforts.  4 Mbps doesn't sound too bad, you should still be able to steam great quality in theory.  But if your single threads are limited to a much lower speed that combined 4 Mbps doesn't help you much.
 
Here's a topic that may help also... the same is true for most speed tests online right now with the obvious exception of TestMy.net.  https://testmy.net/ipb/topic/28902-why-do-my-results-differ-from-speedtestnet-ookla-speed-tests/ -- trust me, your ISP hates me... but more and more are starting to work with people using TestMy.net results.  TMN uses time tested methods that are proven... if your ISP wants to ignore facts then maybe you should shop around a little.   :wink:  Look at the ISP Rank tab on the stats page for your area.  The United Kingdom for example, I'd probably look into VirginMedia.  I ignored the first two because they aren't as popular (see #clients), others may be worth checking out too.
post-2-0-94209000-1408109424_thumb.png
 
... I assume that your experience online feels slow otherwise you wouldn't be complaining to them.  Your provider may or may not listen to my results but you can feel which test is right... your testing to a bunch of locations and ending up with subpar performance.  It's not a coincidence, don't let them convince you that it is.  
 
- Good Luck and Happy Testing,
- Damon - TestMy.net
 

 

post-2-0-33523100-1408109420_thumb.png

post-2-0-51693500-1408109421_thumb.png

post-2-0-85942100-1408109422_thumb.png

post-2-0-67647700-1408109423_thumb.png

Link to comment
Share on other sites

×
×
  • Create New...