• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Testing LAN speed

Knocks

Senior member
I am using a free tool called LAN Speed Test (Lite) recommended on several sites and am getting very inconsistent results (see screenshots below). I didn't even think it was possible to be getting a 1.5 Gbps speed with a regular gigabit network card.

Is the software I'm using not very good, or do you guys think my test setup is incorrect? I am using a shared folder to test, which, if I understand correctly, this setup makes the data travel to the router and then to the shared folder and measures that speed, but perhaps that's not the case.

y3x8tQr.png

FrE85LR.png
 
I am using a free tool called LAN Speed Test (Lite) recommended on several sites and am getting very inconsistent results (see screenshots below). I didn't even think it was possible to be getting a 1.5 Gbps speed with a regular gigabit network card.

Is the software I'm using not very good, or do you guys think my test setup is incorrect? I am using a shared folder to test, which, if I understand correctly, this setup makes the data travel to the router and then to the shared folder and measures that speed, but perhaps that's not the case.

y3x8tQr.png

FrE85LR.png

That's the program that I generally use to measure network speeds and I have always gotten pretty consistent results from it.

I'll have to check the screenshots when I get off work and they aren't blocked.
 
To be close to reality and take into consideration that Network multitasking while you are doing the Test.

I do it with 100MB Packet and Not "1,000,000,000MB" as it indicate on your screen shot.

Also I would make a New empty folder on C:\ and use it for this purpose.




😎
 
You should realize that "writing speed" often means how fast a program in userspace can hand over data to the kernel. That speed is not necessarily related to the network speed. When the application is using UDP, the kernel is allowed to drop packets even without attempting to send them out the machine over the wire.

I guess that many application-writers prefer to use UDP for "speed tests", because they have heard that "TCP has a lot of overhead. And UDP is a lot faster".

The fact that you seem to send 1.5Gbps over 1Gbps link is a clear indication that packets are dropped on the floor, probably even before they leave your machine. I think that if an application-writer allows this kind of behaviour, the application is kinda useless.

Also, I wonder if the application really tries to send a packet of 1 million bytes. Or whether it tries to send 1 million bytes in multiple packets. The first thing just makes no sense. An IP packet can not be larger than 65k bytes. So a UDP packet can not be bigger than 65k either. TCP doesn't even have packets, it appears as a byte-stream to the application.

Of course, this is my opinion, before I have read the manual page of the application. 🙂 Everything could be different.
 
Last edited:
Thanks for everyone's opinions.

Btw, I am using an old Cat5 patch cable. Before I staple a 40 ft cable to the wall, should I just go out and buy a Cat5e or Cat6 cable, or is this is much speed as I'm going to get out of this setup?

Cat6 cables are kind of expensive for a 40-ft cable, but Cat5e is like 5 bucks.
 
Thanks for everyone's opinions.

Btw, I am using an old Cat5 patch cable. Before I staple a 40 ft cable to the wall, should I just go out and buy a Cat5e or Cat6 cable, or is this is much speed as I'm going to get out of this setup?

Cat6 cables are kind of expensive for a 40-ft cable, but Cat5e is like 5 bucks.

Go with a Cat5e for gigabit, and try not to pinch the cable with the staples.

use iperf to test speeds.
 
No a Cat5 cable can't do gigabit fine. A Cat5e cable CAN. Cat 5 is two twisted pair and is fast ethernet only up to 100m. Cat5e, 6, 6a and 7 are all four twiste pair and can do gigabit up to 100m.

A lot of people say Cat5 when they mean 5e. Cat5 hasn't been in wide spread use in around 15 years when 5e started taking over.

Just use cat5e. It'll even handle 10 gigabit on a good installation up to 45m. Anything over 5e is way overkill/future proofing the nines/an installation that actually requires it (harsh alien cross talk environment/lots of EM/TONs of cables laid together. IE Data center, server room, large office building, etc).
 
Yeah, I think $5 is reasonable to upgrade to Cat5e, given the different opinions.

What should I use to run the cable along 3 walls and 2 corners (near the floor)? My FIOS cable is stapled nicely, but I don't have the equipment that the Verizon guy had, and it probably doesn't make sense to buy that stapler just to run one cable.
 
Go to Home Depot/Lowes/Home improvement store. They sell cable staples that you use a hammer with for coax/network cables. Like $4-5 for 100 of them.
 
Back
Top