The general understanding is that DSL isn't hobbled by the number of subscribers like cable service is. That is, no matter how many people jump on DSL at any given moment, their speed should remain constant.
I used to believe that too, but lately it’s beginning to look like there's a hole in that theory. A few weeks ago I started noticing painful sluggishness and dropouts when working on my web site, but usually during evenings and nights. It took some time for me to notice that the response times where significantly better in the mornings.
To test the theory (unscientifically, of course) I ran a few ping tests today starting in the morning and stretching into the evening. Sure enough as the day wore on the ping response times grew longer and longer, and not just by some marginal amounts. There's almost 3 times the difference between the morning and night speeds.
Now I don’t know if this is really a DSL problem or just AT&T throttling bandwidth on home users as they get home and jump online, but the end-result is the same, poor performance.
If a picture is worth a thousand words then this speedtest.net snapshot of a test to Los Angeles, CA should succinctly prove my point. Nice grade AT&T, think I'd rather have my dial-up back.
Give a few Sats
A year later and these issues are still around for AT&T. You should mention that AT&T's own "speed test" website is rigged to make your connection look happy, even when it isn't. The mere presence of such a website tells me they are up to no good. And considering I took a college class that covered a whole semester about AT&T and record number lawsuits covering over a century's worth, it's highly likely you have to fight aggressively with this company to see results.
Comment by JackbNimbled — April 24, 2012 @2:36 am