So lately I have felt things were a little sluggish, so I tested my bandwidth using the speakeasy speedtest (i think this is generally a pretty well thought of test). I opened up a browser remotely on my server to run the tests.

here is what he got:

atlanta 0.3 down 1.3 up
chicago 0.6 down 1.0 up
NYC 1.1 down 2.3 up
San Fran 0.3 down 0.4 up

I tried the same tests from my desktop in the office (cable modem):

atlanta 7 down 4 up
chicago 12 down 5 up
NYC 16 down 4 up
San Fran 7 down 3 up

To summarize, my desktop is 4x faster on the upload side and about 20x faster on the download side. Considering our servers are colocated inside a datacenter with oc48 and oc12 lines coming in, this seems unreal. When I sent these to my programmer he half-jokingly asked if we were hosting on dial-up. Considering we are testing across a range of different networks (by using varying cities), and we did a successful tracert test last week – I am thinking that these poor performance numbers aren’t coming from outside the datacenter, but in it.

I am on a 10mbs port, but I am thinking that my datacenter has limited the bandwidth on that switch. He can do that, right?

What kind of numbers do the rest of you get?

Thanks for sharing....