I have a fiber optic connection with 40mbs up and 40mbs down. I have tried various testing services, such speedtest.net and while my download yields excellent results (often times above 40mbs), my upload seems caught at 20mbs (which was also coindidentally the previous speed).
I called the guy who runs the service and he claims that this is normal for faster connections and that the upload measurements are often not "handleable" by most speed testing services. He then gave me a test of his own to use (presumably testing a direct connection between me and the server with no middle men) and it did yield the proper result.
However, every other speed test, and even real world applications such as torrenting, have all said otherwise. Again, I can easily attain an accurate download, but my upload, even when running several high demand torrents that rack up seeders quick and easily get a 2MBS upload going, always seems to cap my upload speeds just shy of 20mbs, I have to imagine this isnt all a coincidence and I have yet to find something to accurately throughput me to above 20mbs upload speed test or otherwise.
The excuse still is that with other bottlenecks plus the idea I guess that upload suffers from more overhead or something that this is normal, but I guess all the evidence I've seen vs. the one speed test he supplied is not sufficient for me (plus the speedtest.net test was just fine when it was a 20/20 connection, so I find it hard to beleive that it just is not able to give an accurate test of a 40/40 because it's too fast or something).
Does anyone have a reccomendation with an accurate internet broadband test I can fire back with? It's hard to replicate the torrent thing, but I do keep an eye on it whenever I fire up 3-4 torrents at once to see where my download goes and then where my upload goes since it's more than enough people on both ends to test the boundaries of my connection.
I called the guy who runs the service and he claims that this is normal for faster connections and that the upload measurements are often not "handleable" by most speed testing services. He then gave me a test of his own to use (presumably testing a direct connection between me and the server with no middle men) and it did yield the proper result.
However, every other speed test, and even real world applications such as torrenting, have all said otherwise. Again, I can easily attain an accurate download, but my upload, even when running several high demand torrents that rack up seeders quick and easily get a 2MBS upload going, always seems to cap my upload speeds just shy of 20mbs, I have to imagine this isnt all a coincidence and I have yet to find something to accurately throughput me to above 20mbs upload speed test or otherwise.
The excuse still is that with other bottlenecks plus the idea I guess that upload suffers from more overhead or something that this is normal, but I guess all the evidence I've seen vs. the one speed test he supplied is not sufficient for me (plus the speedtest.net test was just fine when it was a 20/20 connection, so I find it hard to beleive that it just is not able to give an accurate test of a 40/40 because it's too fast or something).
Does anyone have a reccomendation with an accurate internet broadband test I can fire back with? It's hard to replicate the torrent thing, but I do keep an eye on it whenever I fire up 3-4 torrents at once to see where my download goes and then where my upload goes since it's more than enough people on both ends to test the boundaries of my connection.