9 replies
by techexec2
Q1. How are the 3 desktops and the 1 laptop connected to the LAN? Wireless B? Wireless G? Cat 5? 100 Mbps? 10 Mbps?
Q2. What operating system are the 3 desktops and 1 laptop running?
Q3. How fast are the CPUs in the server, 3 desktops, and 1 laptop?
-----
It sounds like at least 1 of your PCs is connected via Wireless (only 4 hardwired ports on your router). Any node connected via Wireless G has a theoretical maximum of 54 Mbps but will get less at all times. If Wireless B, only 11 Mbps but will get less at all times. If it has a weak signal, even lower still.
Any other traffic on your wireless, or wired, LAN will adversely affect the test.
by echous
A1: 3 desktops connected via Cat5 @ 100 mbps. 1 laptop via Wireless B (11 mbps, 20ft from router)
A2: 3 desktops running 2000 Professional and laptop running XP Home
A3: 2003 Server is 2.4 ghz, nodes are 933, 750, 732 mhz. Laptop is 1.6 ghz.
Also, switched ports to see if 5 mbps stays with port. No it stayed with 2003 Server. Switched Cat5 cable to see if higher speed stayed with cable. No, it again stayed with 2003 Server.
Finally, event viewer showed following...
Event ID: 7000
Event Source: Service Control Manager
Description:
The Network Load Balancing service failed to start due to the following error: The service cannot be started, either because it is disabled or because it has no enabled devices associated with it. Network Load Balancing is not installed on the system.
by techexec2
I ran some tests of my own and got similarly varied results. It appears that CNET's test is GREATLY affected by CPU speed, and also by browser. IE was 12% faster than FF.
So, it is a misnomer for them to call it a "bandwidth test". It is a measure of throughput using your network, computer hardware, AND your browser software, not just bandwidth.
TEST RESULTS
All of these system are connected to the same LAN router via CAT-5 at 100 Mbps:
SYSTEM 1: WINDOWS XP, AMD Athlon XP 3000+
FF 1.5: 2463, 2374, 2331 Kbps
IE 6: 2691, 2693, 2664 Kbps
SYSTEM 2: FEDORA CORE 3 LINUX, AMD Athlon 64 3000+
FF 1.5: 2691, 2696, 2695 Kbps
SYSTEM 3: FEDORA CORE 5 LINUX, AMD K6-2/500 MHz
FF 1.5: 879, 892, 870 Kbps
by echous
I don't know where to go from here. Is there a better test site? Is anything broken? Even at "2 mbps", IE performance isn't typically a problem, although the "5 mbps" server surfing response is definately nice.
Next step for me is to Cat5 the laptop to see how it's higher speed responds to a 100 mbps network vs the 11 mbps wireless.
Thanks for help.
by techexec2
My test results, and yours, showed that there is a strong correlation between CPU speed and the CNET Bandwidth Speed Test. That is, CPU speed really matters when it comes to getting the most speed out of web browsing on a high-speed Internet connection.
So, you must upgrade your four slower systems (3 desktops, laptop) to higher speed (go for around 3.0 GHz...good price point) in order to get the most out of your 5.0 Mbps cable modem with them.
Even if you don't upgrade them, 2 Mbps is pretty darn good. Just a short few years ago, "high speed Internet" by cable modem was 300 Kbps (0.300 Mbps) and most people had 28.8 Kbps modems (0.028 Mbps).
Whine whine whine!!!
You're doing very well. Take care.
:^0
B-)
by techexec2
One computer on his LAN gets 5 Mbps. The others are only getting 2 Mbps.
Read the entire thread and you'll see what is really going on here.
Merry Christmas to you...
P.S. Good information about coaxial cable and splitters. It just doesn't apply to this guy's case.
by jdclyde
it is good information for anyone that is adding more outlets for their TV's, but you are correct that this has no bearing on this case.
Once it is handed off the the access point, the question is where do the PC's get their DNS from? Do they all get it from the internet provider, or from the server? If they have to poll the server and THEN go out, that is a possible hit.
Are they on-board NIC's in the PC's? Again, a performance hit.
by jtumin
This discussion is closed to new replies.
Trending Networks posts
Sort by Trending