Reliable speed test?

Mopower

Gawd
Joined
Feb 13, 2006
Messages
519
I have 15 Mbps down and 1 Mbps up DSL. Why is there only one server I do a speed test from shows me the correct speed? My ISP speed test shows 7 Mbps down. Speakeasy shows 3. The only one that shows what I am suppose to be getting is if I choose the Chicago server on Speedtest.net. Any place else on Speedtest.net shows 2-4 Mbps and ping times over 100ms. The Chicago server I get 14 Mbps down and pings in the 50's. Chicago would be the closest but that doesn't explain why my ISP's speed test is so far off. Any ideas?
 
There are a lot of variables on DSL. Depending on where you are in the network or the wiring in your house 2-5 Mbps with around 100ms ping may end up being as good as it gets.
 
There are a lot of variables on DSL. Depending on where you are in the network or the wiring in your house 2-5 Mbps with around 100ms ping may end up being as good as it gets.


Then why does that one server say the speed I'm suppose to be getting but the others vary wildly? I know how DSL works I install it for a living. I get the same results doing speedtests at the CO.
 
You've just discovered how the internet works. :) It's not 100% consistent across the world. Hardware, distance, load, remote connection and 1000 other variables in between effect transfer speeds and response times.
 
Just wondering if there is something wrong since every speedtest except the chicago server on speedtest.net is incorrect.
 
Just wondering if there is something wrong since every speedtest except the chicago server on speedtest.net is incorrect.

It's not that they are incorrect, it's just the variables in between you and what you are downloading are optimal where you can max your line.
 
Do tests when network load is low, like early morning on weekends and see if problem persists.
 
The Chicago server at speedtest.net is the only one I've found to be realiable 'most of the time' for me personally. I'm also within range. But yes, late in the wee hours between 1-5am is the best time.
 
your connection speed is dependent on every single link in the chain that the traffic takes to get from the server to your PC.

the only way every test server would show "accurate" numbers as you call it, would be if the only bottleneck was the link between your house and the first node on your ISP's network. but if the bottleneck changes places when you go thru different virtual paths to get to different servers, the speed test numbers change.

none of them are "incorrect". they reflect a different path taken to get to a different server. does it take you the same amount of time to travel 10 miles on a highway and 10 miles in the city? no. not all paths are equal even if they are the same distance.
 
Also depends on where the fiber that provides the backbone for your DSL line comes from. Most likely your DSL is coming out of a chicago based fiber backbone, and thus you get the best connection because that's where your connection actually originates.

Personally I always test NYC because it is a large hub on the east cost and more then 100miles from me.
 
Last edited:
Back
Top