Speed Tests

[#1]

@Ian06  said.

---

@Ian06 wrote:

@greygit1   The traceroute isn't going via the VPN because as @TheRoadCrew pointed out the VPN runs from within the Firefox browser, so I believe the VPN is only used when you are in/using the browser - anything done outside the browser doesn't go through the VPN I guess. 

Yes, Undestood... i.e. the "command line" or in Linux " Terminal" instruction traceroute would not go through the VPN.

When @shutter ran the Speed Test with Ookla I believe he would have done it from within the browser and hence could use the VPN.

Yes, again... the speed test was done with the VPN ON and also with the VPN OFF 

The Speed test results showed that when the VPN was ON... the speed test result was at least TWICE as fast as when not using the VPN.

---

[#2]

@Ian06 

---

@Ian06 wrote:

---

Anyone know if urban VPN is reputable, safe to download and what the trade off is for a "Free" VPN?

---

I have been using Urban VPN for a couple of years now..... No problems... 

TRADE OFF  ?   IS doubling up the speed a trade off   ? ?   

[#3]

To all of you who have contributed to this thread....

    Thank you..

                         for your interest.... and for your opinions and postings

It does not answer the question, but has given an insight into how complex the internet and how it works, it really is.

Although the "general thinking" seems to be that the further the traffic/packets have to travel  the longer it will take to get there... as shown by the PING value on the results posted... Other things mentioned was encryption, adding a time delay along the journey...but neither of them seem to be working against the actual results showing a doubling and sometimes trebling of my "standard" speed when not using a VPN

As I do not need the excessive speeds published by the 4 main ISP`s.. and I am happy with what I get.. as it "works for me" . this has just been an exercise in trying to find out WHY something works, the way it does !..

and I`m still baffled.

Never Mind. !.... but Thank you  to you all..

Have a Happy New Year.

cheers

[#4]

Hello All !....

As it was mentioned, and confirmed.... that the Urban free VPN provided by Firefox Add-ons... was only working through/on the browser. and there was mention that it could be used on Windows for all connections...(but not supported for Linux )  I downloaded the windows version, to see if it would run within WINE 8.0.2 .

It went through the usual install procedure, and showed the icon on the lower task bar ( Linux version " lower Panel ") . on clicking that. it showed a gui panel, and I selected Germany... it then started trying to connect...but never did...

Googled for a Free VPN for Linux mint... and came up with PROTON .  downloaded the package, installed it, set my user name and password and YAY  ! it works.

so here is some very interesting comparisons of speed tests  with and without the PROTON VPN in operation.

First one is using the "FAST" speed test site.... note the Proton stating "you are disconnected".

And now..... a real shocker...    

WITH THE PROTON VPN connnected to the United States of America !

Kinda knocks much of what was considered regarding the use of the freeby vpn on the browser....

To check that it was not a FREAKY result.... I then tried the OOKLA speed test...

here it is with NO VPN connection...

and here again.... shock result.... with PROTON VPN connected...

Double checking... with NO VPN connected on the FAST site

and then connecting to the NETHERLANDS

So there you have it....... Make of it what you wil..... it still seems a bit weird to me ! ! ...

Have a good evening.

cheers

[#5]

---

---

@TheRoadCrew wrote

[...]

---

@Ian06 Ian06 wrote:

Beyond my experience - so may be way off beam....  perhaps it's time for me to keep quiet..!!

---

I'm beginning to feel that way too 🙄

---

@Ian06 my apologies if my comment above seemed aimed at you; it was intended to read as I, too, was beyond my experience and perhaps off beam.

Returning to the thread now I can see it could be easily misinterpreted.

[#6]

"Make of it what you wil....."

The latencies and throughput are both being reported via a web interface served up the web browser? And (therefore) must be using HTTP 'ping' (which is a different beast to commandline ping in *ix and Windoze).

IIRC Proton operates out of Switzerland (started off as connected to CERN)

https://protonvpn.com/

"

VPN Accelerator

VPN Accelerator is a set of technologies unique to Proton VPN that can increase your VPN speeds by up to 400%. We created VPN Accelerator to dramatically increase speed and performance by overcoming CPU limitations, using advanced networking techniques, and redesigning VPN protocols themselves.."

(just looking at the proton VPN website)

[#7]

---

@TheRoadCrew wrote:

---

---

---

@Ian06 my apologies if my comment above seemed aimed at you; it was intended to read as I, too, was beyond my experience and perhaps off beam.

---

It didn't seem to be aimed at anyone, I read it the way you intended, no offence taken.

[#8]

---

@shutter wrote:

here is some very interesting comparisons of speed tests  with and without the PROTON VPN in operation.

......

So there you have it....... Make of it what you will..... it still seems a bit weird to me ! ! ...

 It would still be interesting to see the traceroute (to the speed test servers) with and without PROTON VPN to compare the routes and speeds....  I wasn't sure how to find the address of the speedtest servers but after a little effort I think I found the Oxford University, Abingdon server they use 129.67.206.81    ---    far-80g.oxide.ox.ac.uk

The route via the VPN network must either be a lot shorter, compression is used effectively to reduce the size of the payload or the internal connectivity of the VPN is quite different to what we are thinking.

Is there any configuration in PROTON VPN that might switch compression on or off?

[#9]

@Ian06 

OK.... traceroute  -     traceroute far-80g.oxide.ox.ac.uk

NO VPN

Trace route with VPN on USA

Speed test      NO VPN . using ABINGDON  SERVER

Speed test  using ABINGDON SERVER      VIA   USA   VPN ON

The above VPN was the PROTON vpn installed as previously mentioned.

[#10]

Thanks @shutter I will try and work my way through these and see if it tells me anything.... but it clearly shows the impact of the VPN in terms of routing...  I don't really know it will help me much but the last few hops of the traceroutes are missing. I may be some time....

Anyone else familiar with this sort of output is welcome to give us the benefit of their knowledge - this is really a learning exercise for me!!

My guess still is that the VPN is doing a good job of compressing the data that is exchanged for the speed test and hence showing a much better test result

[#11]

This is my best efforts looking at the traces (sorry if it's disappointing, but I'm not a network engineer...)

VPN Route:
Hop IP Adrs         Comments.....
1.    10.96.0.1           Private IP adrs - assume it is the VPN internet gateway
2.    31.13.189.225   Router in New York - assume the public IP of your VPN
3.   146.70.0.221     Secaucus, New Jersey, United States - assume an internet router
4.   62.115.183.74    Jersey City, New Jersey, United State - assume an internet router
5.   62.115.136.46    New York City, New York, United States - assume an internet router
6.   62.115.113.21    London, England, United Kingdom
7.   62.115.122.189  Hendon, England, United Kingdom
......
11. 146.97.37.194  Hayes, England, United Kingdom -
12. 193.63.108.70 London, England, United Kingdom
13. 193.63.109.42  Oxford, England, United Kingdom
...... rest of trace is missing, but we are close to the speed test server

non-VPN Route:
Hop IP Adrs         Comments.....
1.     192.168.1.1        Private IP Address - shutter's internet router/gateway
..... some hidden/non-responding routers in Three's network?
6.     172.16.216.93    Private IP Address - presumably inside Three's network
.....
8.     172.16.195.146  Private IP Address - presumably inside Three's network
..... (I'm not sure why shutter's public IP address doesn't seem to have show up here somewhere)
10.   195.66.224.15  London, England, United Kingdom - assume an internet router
... then it seems to take similar router addresses to Oxford University servers

The round trip times shown to each of the routers aren't very different. My guess still is that the VPN application is successfully compressing the speed test data across the VPN tunnel so the total data is travelling more quickly over the slower parts of the connection. Going much quicker to and from New York over the VPN (because is has been compress to a much smaller size) than the uncompressed data is able to travel through Three's network.

Across the other hops outside of Three's network and outside the VPN the speed of transmission is probably very quick and not constrained by the 4G network connection....

[#12]

@Ian06  Thank you very much for all your detective work on the route and stuff...

I suppose that what you have found out about PROTON VPN.. and the way it (seems) to work, could also apply to the URBAN FREE VPN , as the speed results are very similar, (notwithstanding the different distances between Frankfurt and Abingdon,... and the USA and Abingdon. )

So... the "press report" on whether a VPN can/does increase speed, seems to be wrong.... but this is only from an un-educated ( i.t. ) observer, of results obtained. using different speed testers, and different vpn`s.

I think that at the moment, unless anyone else can offere any significant info.... we can consider this a closed thread....

Thank you all for your interest and inputs.

Have a good weekend

Cheers

[#13]

---

@shutter wrote:

@TheRoadCrew  Yes. it does have a page fo QoS settings... will PM you a copy of it ....

---

Thanks; looking at the screenshot you sent and the online manual there are no pre-defined QoS settings.

if you normally leave the router on 24/7 the manual has instructions on how to schedule regular reboots which might improve performance (as might a firmware upgrade if available).

As an aside....

I gave Proton VPN a spin on both laptop and Android phone.

Testing with my VirginMedia connection without vpn and with vpn (first Netherlands server then US server) i got these results: 

Speedtests from within browser were slightly different but vpn results were both slower.

My 4G download speeds on mobile data (GiffGaff) are generally lower than yours; Proton VPN gave me a slight increase using a Netherlands server but a US server was slower than non-vpn use.

Whatever VPNs are doing for your connection isn't being replicated on mine (not that I know why).

[#14]

@TheRoadCrew  OK>... thanks for your input..... strange how there doesn`t seem to be a vast difference between vpn and no vpn speeds on your setup.

I found an interesting speed test site that has a lot more information available about the route the speed test does whether vpn or no vpn.

Check out  http://www.dslreports.com/speedtest

You can select your connection type  and on the "More" it shows 4G.. so I selected that for my tests...

all results using their tester gave similar speeds and differences to OOKLA and FAST... however, if you click on "Reports and Share" you get a lot more info ...

scroll down to the bottom for some more "Tools"..

Have fun

( Note... this did not work using the Installed version of PROTON VPN..... said there was something on the browser blocking it  --- but it did work on the Firefox UrbanVPN )

[#15]

---

@shutter wrote:

[...]

Check out  http://www.dslreports.com/speedtest

You can select your connection type  and on the "More" it shows 4G.. so I selected that for my tests...

all results using their tester gave similar speeds and differences to OOKLA and FAST... however, if you click on "Reports and Share" you get a lot more info ...

[...]

( Note... this did not work using the Installed version of PROTON VPN..... said there was something on the browser blocking it  --- but it did work on the Firefox UrbanVPN )

---

Gave it a quick run on laptop with Proton VPN (Japan and Netherlands servers) and without VPN: