If you’re new to Canopy and just getting started like me, here’s a great site for testing your connection speed when your doing some pre-deployment bandwidth configuration testing:
http://www.speakeasy.net/speedtest/
It’s one of the better ones I’ve seen and it’s pretty accurate. After NUMEROUS bandwidth configurations, the speeds seem to be pretty congruent with my settings. Plus, it’s set up as a speedometer so you can actually see the little bursts of bandwidth. Oh, and best of all…it’s free! 8)
If someone has a better site, please post it here.
These sites are useless. All you are doing is testing the speed and quality of the server/s and paths through the Internet to you. I just tried it and got a report of 3.5Mbps on a GigE link with over 100Mbps of bandwidth available (bidirectionally).
Never, and I mean never, let your customers use these things gladly, they will draw all sorts of wacky conclusions about your company, your subscription to bandwidth ratio, etc. etc.
Shaman, then how did you initially test your bandwidth before deployment?
Running the link test in my SM is constantly giving me the same numbers no matter what the configuration. Any ideas would be appreciated, because I’m trying to figure out the right configuration for 3 different “tiers” of service. My equipment is 5.2 GHz Advantage for both the SM and AP.
Thanks in advance!
Is the configuration source in the AP set to the AP or the SM?
Using the AP forces all SM’s to the speed set in the AP. Using the SM allows you to adjust speed per subscriber.
It’s set to the SM. I actually picked up that info from your older thread http://motorola.canopywireless.com/support/community/viewtopic.php?t=611 
I’ll provide more info for you in the “bandwidth tests” thread.
As always…thanks Jerry!
Set up a file on you web server that you can download.
Trust me here, those speed tests are useless. Especially the ones run by ISPs, do you really trust them to show you that you have a faster connection than they are offering for sale?
Shaman666 wrote: Trust me here, those speed tests are useless. Especially the ones run by ISPs, do you really trust them to show you that you have a faster connection than they are offering for sale?
No, and that would probably explain why I got a different speed at almost every site.
But what threw me off with the site I listed above was how close the numbers were to my Sustained Data Rates. When I configured my SM for 6Mb down and 1 Mb up, the figures on this site weren't exact but pretty close. Same for 3.5 down/1.5 up and 1.5 down/512 up.
I tried all of the above with the Downlink Data % set to 90%, 85%, 80%, and 75%. And my best results were when the Downlink Data % was set to 75% in my AP.
This is what made me think I was on the right track, but I guess I'm just taking the wrong approach.
I guess I better try what you said instead. Thanks for the feedback Shaman! :)
We use Speakeasy.net or another good one has been Testmy.net,
Testmy.net will sell or trade a test program you can put on your server set up however you like
Cool 8) thanks!
we’re just in the process of starting out with canopy but also have around 3,500 cable, dsl, and FTTP customes and have had pretty good results using speakeasy. it’s not perfect but seems to have pretty consistant results.
As far as the “accuracy” of any speed test on the Internet at large, what’s the point of worrying over it? Customers, in general, don’t care if you tell them “That’s not actually accurate”.
I agree with the original poster that Speakeasy is the best-of-breed, even if it isn’t technically perfect.
You can license the “speed test” software that Speakeasy uses for a nominal yearly charge, and run it on a server at your head-end btw.
ais3101 wrote:
Running the link test in my SM is constantly giving me the same numbers no matter what the configuration.
That's exaclty what is supposed to happen.
The link test option in the APs and SMs test the raw RF throughput through the link. It does not take into account any OSI layer 3 bandwidth limiting you may have configured.
ahull wrote:
That's exaclty what is supposed to happen.
The link test option in the APs and SMs test the raw RF throughput through the link. It does not take into account any OSI layer 3 bandwidth limiting you may have configured.
Ok, that's what I figured, but then again I wasn't sure.