BT SPEEDS BEING THROTTLED ? ADVISE

Soldato
Joined
5 Dec 2006
Posts
15,370
Thats like me saying Virgin aren't providing me 50Mb because HillBilly Joe's speedtest server in the USA isn't giving me 50Mb. I can't complain to Virgin about their routing and partnering links, and their partners partner links, and their partners partners partners links etc etc etc (see where I'm going?)

I'm not talking about HillBilly Joes speedtest server though.

I'm talking about speedtest servers provided by high tier internet backbone providers. If you cant even get max throughput to major data centres in NL and other major IXPs then your connection to the "internet" is quite frankly pointless I'm afraid.

You can continue believing that your "internet" is amazing just because you're getting advertised speed to your local node or whatever. I guess it makes you happy to finally see a figure of their advertised internet speed - even though the "internet" is not even located at the same endpoint their running the test to!

Personally I run tests to where the internet actually is. That is all over the world.
 
Last edited:
Caporegime
Joined
17 Jul 2010
Posts
25,714
No sorry. You are mistaken. Every speed test is accurate. Every single one.

When you download a file from NL and you get 100Mb/s, it is accurate. When you download a file from Scotland and you get 20Mb/s on the same connection it is still accurate. Anywhere you download a file from is accurate speedwise. It won't say you're throughput is 100mb/s when you're actually downloading at 20mb/s now will it?

What I'm saying is a speed test to your local ISP's local speed test server is absolutely pointless, not inaccurate. It's got nothing do do with inaccuracy, it's to do with pointlessness because the internet is not located at the same endpoint as BT's speed test server.

In fact internet routing goes anywhere between 7-15 levels deeper than the first hop to your ISP. And even deeper if your ISP has crap routing.

I would disagree that every speed test is accurate. Speedtest.net give no indication of the capabilities or load on any server you choose. If one particular server is getting absolutely hammered and has insufficient bandwidth to serve all these testers then you're being given a false result compared to how your speed impacts your other internet activities.

My ISP is Sky. I still get great results on the BT Wholesale site but often get terrible ones on Speedtest.net.
 
Man of Honour
Joined
13 Oct 2006
Posts
91,052
Very rare I don't get a true reflection of my potential speeds from speedtest.net - the odd time it will connect me to a ropey server but usually for anything upto 100Mbit its been accurate - if I don't get what I think to be a correct speed from one speed test I'll cross reference with others however.

BT provide you a speed, if the speedtest is unfiltered or prioritised, surely the service is still performing to that level? All traffic cannot be high priority, there has to be allowances for real time or critical protocols.

Where do you draw the line though? if the BT Wholesale test is running full speed but all proper internet traffic is running at like 0.5MBit/s that is hardly living upto what you'd expect from being sold a speed.
 
Last edited:
Soldato
Joined
5 Nov 2011
Posts
5,361
Location
Derbyshire
You can continue believing that your "internet" is amazing just because you're getting advertised speed to your local node or whatever. I guess it makes you happy to finally see a figure of their advertised internet speed - even though the "internet" is not even located at the same endpoint their running the test to!

Personally I run tests to where the internet actually is. That is all over the world.

However your ISP is only interested in what they have control over. Yeah there's arguments their partner links may not be great but BT's network is their responsibility.

Where do you draw the line though? if the BT Wholesale test is running full speed but all proper internet traffic is running at like 0.5MBit/s that is hardly living upto what you'd expect from being sold a speed.

OP seems to be describing the issue at peak times though so is probably contention orientated with traffic shaping doing what it is designed to do. Give a result that proves the actual line can do the correct speed and prioritise their chosen golden protocols, more than likely BT vision stuff.
 
Back
Top Bottom