
ET
May 26, 2003, 10:57 PM
Post #5 of 19
(4623 views)
Shortcut
|
Re: [RexC] Internet connections / checking speeds
|
Can't Post | Private Reply
|
RexC writes: I can't advise you for a connection in your area, but click on the link above to check your actual connection speeds. This test site is in Texas. You can run up to 3 free tests per hour. A speed test takes about a minute. You can test dial up, and broadband speeds, including T1. Speeds do vary sometimes when using cable connections. I have a 256 KPS cable connection here in Cuernavaca. I usually connect at 232 - 244 KPS, only once was it as slow as 200 KPS.... ....I know this won't help you with your decision, but it might show if T1 is worthwhile in your area. In the US full T1 connections are still running between $300 and $700 USD per month, depending on the tier of the provider, the type of service level agreement you sign, and the actual use of the connection. As there's less competing providers in Mexico, I suspect the prices are going to be higher. Considering the original poster is fretting in another forum about paying $300-350 MXP for housing (the time increment wasn't clear) either they've got a very gearhead set of priorities, or there's confusion over what kind of connection they're (a) currently using ("I am using a friend's T1 connection via video cable...." ok....) and (b) looking for. With any web-based "connection" speed test, including the one that you referenced, it's important to note is what you're actually measuring is the speed for the entire route between your computer and the computer hosting the test software. As an example, running a traceroute between the machine I'm punching on right now, and the bandwidthplace.com machine yields the following routing and hop times: 1 [edited for privacy] 30.009 ms 1.695 ms 8.963 ms 2 [edited for privacy] 13.1 ms 15.092 ms 14.004 ms 3 [edited for privacy] 16.164 ms 14.765 ms 12 ms 4 [edited for privacy] 12.166 ms 12.08 ms 13.166 ms 5 [edited for privacy] 13.857 ms 14.572 ms 13.609 ms 6 peer-01.palo.twtelecom.net (198.32.176.111) 13.758 ms 13.843 ms 14.07 ms 7 dist-01-so-3-2-0-0.okld.twtelecom.net (66.192.250.37) 16.057 ms 17.316 ms 15.319 ms 8 dist-02-ge-2-3-0-0.okld.twtelecom.net (168.215.55.86) 15.098 ms 16.071 ms 15.592 ms 9 dist-01-so-0-0-0-0.snfr.twtelecom.net (168.215.55.73) 17.188 ms 16.115 ms 15.384 ms 10 10.99.99.1 (10.99.99.1) 26.18 ms 25.593 ms 26.273 ms 11 core-01-so-2-2-0-0.dlfw.twtelecom.net (168.215.53.89) 60.396 ms 62.471 ms 59.847 ms 12 dist-01-so-0-0-0-0.ausu.twtelecom.net (168.215.53.94) 65.179 ms 67.577 ms 64.49 ms 13 dist-02-ge-3-3-0-0.ausu.twtelecom.net (168.215.53.166) 64.224 ms 64.589 ms 65.026 ms 14 dist-01-so-0-0-0-0.snan.twtelecom.net (168.215.53.105) 66.214 ms 66.703 ms 66.175 ms 15 64-132-228-250.gen.twtelecom.net (64.132.228.250) 68.101 ms 66.005 ms 65.597 ms 16 64-132-228-26.gen.twtelecom.net (64.132.228.26) 58.496 ms 59.31 ms 58.068 ms 17 vl130.core1.sat.rackspace.com (64.39.2.33) 61.661 ms 58.174 ms 58.106 ms 18 vl901.aggr1.sat.rackspace.com (64.39.2.66) 61.1 ms 58.491 ms 60.178 ms 19 bandwidthplace.com (209.61.187.19) 58.954 ms 62.724 ms 61.056 ms What this is showing that between my machine and the bandwidthplace.com speed test server, the traceroute test packets, and presumably the speed test packets and everything else pass through a total of 18 intermediary routers. If even one of these routers is congested or experiencing problems, a lower level of performance is reported. At the same time, if during your normal use your information never passes through the same problem router, you'll never experience the slowdown, and conversely if most of your information flows through a route that's more congested you may always get poorer performance than the test indicates. To counter this problem it's recommended that you choose a test server with the lowest routing hop count, which generally translates to one in close geographical proximity to you; that you test your connection multiple times at different times of the day and days of the week and that you only consider the average of the test results; and that when you test your results you also run and examine a traceroute to the test server to identify any changes or anomalies in routing.
|