said by Webslinger:No, I'm saying a couple of things. The first, and most important one is, if my neighbour on Rogers is getting significantly faster throughput (on a slower speed tier than I am) to pretty close servers, then the issue can't be blamed on node congestion, which seems to be a common excuse in other threads.
The second thing I'm saying is that if I'm getting 50% (or less) of what I should be getting from a close server, and if my Rogers neighbour (same test on a slower speed tier) is getting close to 10 MB/s more than I am at the same time, then I'm pretty sure there's a problem. And I don't feel that problem can be fully explained by routing either.
This comes back to something I suggested to TSI Marc about 15 month ago (IIRC).
At TSI's expense, install a both TSI line and an 'incumbent' service at the same address (ie. within each area served by a cable POI or telco CO, & Stinger area & 7330 area). Install identical modems and identical computers directly attached to each modem. This would be dedicated for TSI testing purposes. Run identical scripts on each computer 24/7 to test the connection quantitatively for all modem stats and different types of user behaviour (torrents, etc....) all controlled by scripts.....no human interaction required.
Send the results (every 10 minutes or so) to a database @ TSI and run analytics on it.
This should be done for cable and xDSL, wherever it can be done. Only in this way can TSI/CNOC ever hope to understand what, if anything, the incumbents are doing to further screw with them.
