Asymmetrical performance on FIOS Gigabit
jdri
Enthusiast - Level 1

My G1100 is hard-wired from the main input black box (basement) by the technician when upgrading to gigabit.

I have an ASUS RT 5300 on a network port acting as my primary router with the G1100 WiFi off. 

Validation of speed shows: 

Download 419Mbps Upload 706Mbps via SPEEDTEST

Is this a standard result?  

If not, what is driving the issue? 

Any specific guidance is appreciated. 

0 Likes
1 Solution

Correct answers
Re: Asymmetrical performance on FIOS Gigabit
lasagna
Community Leader
Community Leader

Both responders here are correct (and slightly incorrect at the same time).

The Vz test only senses traffic within the Vz network (and can do both between your device and their test node and between your ONT and the test node).   Basically it only tells you want Vz has you provisioned at and should always report at or near your contracted bandwidth (keeping in mind the "up to" nature of the gigabit offering -- other speeds are "as stated").

Tests to other networks (such as speedtest.net, etc.) are subject to many factors including the interconnection congestion if the server is on a different ISP, the speed of the server's internet connection, and traffic in general.   Keep in mind some servers might also be operating on asymmetric connections which are typically higher speed heading toward the server and slower coming back -- which is why your "upload" (the remote server's download) may be faster than your download (the other's server's upload).

There's also the "conspiracy theory" aspect of some test sites -- in that tests from Verizon's network to servers sponsored and run by another ISP might purposely look "slower" so as to cast the performance of the other ISP's network in a better light.  No proof and I'm not naming any names (and thus any ISP who takes offense should probably look at themselves in the mirror).    This could also just be that such larger ISP provider servers are more frequently the target of a test and if multiple people test simultaneously, you are competing for bandwidth (which is why you need to not be transferring any other data one your own machine or network while testing).  I always test against several different server endpoints -- some of which will always consistently max out at a certain speed and others which will sometimes show higher speeds just to get a sense of the true reality.

And lastly, as one of the posters mentioned, speed tests that are browser based are somewhat constrained by the mechanics of the browser itself.  If you have a higher speed connection, you should look into using one of the test network apps to run your test which will tune it's behavior and not be constrained by the browser limitations in order to get a true sense of your actual speed.

View solution in original post

Re: Asymmetrical performance on FIOS Gigabit
jdri
Enthusiast - Level 1

Verizon test shows 

943.5 Download 938.89 Upload

0 Likes
Re: Asymmetrical performance on FIOS Gigabit
JustinG1
Community Leader
Community Leader

Those tests look fine.

Regarding your question about performance, FiOS "Gigabit" is marketing hype. You won't see a full 1,000 Mbps. (If you read the fine print on the Gigabit page, it states connections up too 940 Mbps down and 880 Mbps up). So your connection is actually getting more than whats advertised. 

The Verizon speed test is your true speed test. All others from 3rd parties can't be trusted to be a fully reliable test, as there are many factors effecting traffic.

Personally, I'm starting to not trust speedtest.net as much anymore, especially for gigabit speed connections. At my work we recently had a 1 Gbps fiber link installed (not from Verizon, but from a commercial ISP Cogent), and speedtest.net showed a mere 200 Mbps (when the link was in fact true 1 Gbps). 

Best of luck. 

Re: Asymmetrical performance on FIOS Gigabit
tvxnt
Newbie

Sorry, have to comment here to get rid of some bad info.

Correct, Verizon advertises 940Mbps/880 down/up. 

But Verizon speed test is not the best solution. 

Ookla's speedtest.net caps out at 3Gbps due to browser limitation. Your best bet is to download speedtest.net's desktop app and test that way. 

You're trying to check the connection from endpoint to endpoint, not from what Verizon is sending to you. There's a load of issues that can happen between what Verizon's sending and what you're getting. 

0 Likes
Re: Asymmetrical performance on FIOS Gigabit
CRobGauth
Community Leader
Community Leader

@tvxnt wrote:

Sorry, have to comment here to get rid of some bad info.

Correct, Verizon advertises 940Mbps/880 down/up. 

But Verizon speed test is not the best solution. 

Ookla's speedtest.net caps out at 3Gbps due to browser limitation. Your best bet is to download speedtest.net's desktop app and test that way. 

You're trying to check the connection from endpoint to endpoint, not from what Verizon is sending to you. There's a load of issues that can happen between what Verizon's sending and what you're getting. 


Verizon Speedtest is the test to prove you are getting what you paid for.

All you are promised (by Verizon and any other ISP) is the speed to access thier network. That is what the Verizon speedtest checks. Anything past their network and there could be a variety of issues a number of which are not under the control of your ISP.

Re: Asymmetrical performance on FIOS Gigabit
lasagna
Community Leader
Community Leader

Both responders here are correct (and slightly incorrect at the same time).

The Vz test only senses traffic within the Vz network (and can do both between your device and their test node and between your ONT and the test node).   Basically it only tells you want Vz has you provisioned at and should always report at or near your contracted bandwidth (keeping in mind the "up to" nature of the gigabit offering -- other speeds are "as stated").

Tests to other networks (such as speedtest.net, etc.) are subject to many factors including the interconnection congestion if the server is on a different ISP, the speed of the server's internet connection, and traffic in general.   Keep in mind some servers might also be operating on asymmetric connections which are typically higher speed heading toward the server and slower coming back -- which is why your "upload" (the remote server's download) may be faster than your download (the other's server's upload).

There's also the "conspiracy theory" aspect of some test sites -- in that tests from Verizon's network to servers sponsored and run by another ISP might purposely look "slower" so as to cast the performance of the other ISP's network in a better light.  No proof and I'm not naming any names (and thus any ISP who takes offense should probably look at themselves in the mirror).    This could also just be that such larger ISP provider servers are more frequently the target of a test and if multiple people test simultaneously, you are competing for bandwidth (which is why you need to not be transferring any other data one your own machine or network while testing).  I always test against several different server endpoints -- some of which will always consistently max out at a certain speed and others which will sometimes show higher speeds just to get a sense of the true reality.

And lastly, as one of the posters mentioned, speed tests that are browser based are somewhat constrained by the mechanics of the browser itself.  If you have a higher speed connection, you should look into using one of the test network apps to run your test which will tune it's behavior and not be constrained by the browser limitations in order to get a true sense of your actual speed.