WOW This makes no sense!!

[anox195]

Ok anyways, I have a 6/384Kbps connection with Comcast right now in my home. I have a Linksys router that has a maximum output speed to be 100Mbps and I have a ethernet Cat 6 cable going to a Linksys 5-port hub, who's output is only up to 10Mbps ((**it says anyways**)). So I have a ethernet Cat 6 cable going from the hub to my computer, Linksys 10/100 Ethernet NIC. And these are the results coming up right now.

::::::::::.. Download Stats ..::::::::::

Download Connection is:: 12574 Kbps about 12.6 Mbps (tested with 12160 kB)

Download Speed is:: 1535 kB/s

Tested From:: https://testmy.net/ (Main)

Test Time:: 2007/10/03 - 3:40pm

Bottom Line:: 219X faster than 56K 1MB Download in 0.67 sec

Tested from a 12160 kB file and took 7.922 seconds to complete

Download Diagnosis:: Awesome! 20% + : 112.79 % faster than the average for host (Comcast.net)

D-Validation Link:: https://testmy.net/stats/id-R2QUV4XG5

User Agent:: Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 5.1; .NET CLR 1.0.3705; .NET CLR 1.1.4322; .NET CLR 2.0.50727; Media Center PC 4.0; .NET CLR 3.0.04506.30)

How am I going 2.5Mbps above the maximum output? Without the hub I can get up to 16Mbps, thanks to Comcast's Powerboost technology.

But can anyone explain the 2.5Mbps above maximum output on the hub?

Does Linksys just kinda "*guess*" maximum output and not actually test their hardware?

[nanobot]

It could be the fact it is only going to the first router and not toyour computer i dunno but its a possibility...

[dlewis23]

Its just a burst in speed. Your not really getting that.

I had wondered how that could of happened as well upto about 2 weeks ago when I saw the same thing happen on one of my servers with a 10mbps port speed that would go to 13mbps, and was told that it can burst to that speed for a few seconds. Its nothing to worry about.

I wouldn't keep it on that hub tho.

[anox195]

Well, according to the Comcast Tech I talked to over the phone about powerboost technology, he says that when the surrounding network has capacity that isn't being used then I will get a increase up to 14-16mbps download once I initiate a file transfer of any kind. Although he said that powerboost only affects download speeds, he said it does not affect upload in anyway. Im assuming they increased upload throughput to 1.5mbps because I obtain about 1.7 when testing, no where near 384kbps. But anyways to get back to the point, technically by him, I am getting those speeds, just for a short time, only when there is no heavy traffic on the network. I live in a small town and haven't noticed a download without this powerboost yet.

:::.. Upload Stats ..:::

Upload Connection is:: 1724 Kbps about 1.7 Mbps (tested with 2992 kB)

Upload Speed is:: 210 kB/s

Tested From:: https://testmy.net/ (Main)

Test Time:: 2007/10/03 - 7:38pm

Bottom Line:: 30X faster than 56K 1MB Upload in 4.88 sec

Tested from a 2992 kB file and took 14.219 seconds to complete

Upload Diagnosis:: Awesome! 20% + : 208.96 % faster than the average for host (comcast.net)

U-Validation Link:: https://testmy.net/stats/id-EZ1MVRW0G

User Agent:: Mozilla/4.0 (compatible; MSIE 7.0; Windows NT 5.1; .NET CLR 1.0.3705; .NET CLR 1.1.4322; .NET CLR 2.0.50727; Media Center PC 4.0; .NET CLR 3.0.04506.30) [!]

So technically occording to him I am obtaining those speeds, only for a short period of time when the network will allow it.

I just don't understand how the hub is allowing 25% more speed than its maximum output. Mindboggling 

[Swimmer]

I dont think that the hub is allowing those speeds.. my guess is that the router is buffering the IP packets for you there by creating a caching effect..  The best way to see if this is happening would be to run a throughput test on the LAN itself.  Download something like iperf, http://dast.nlanr.net/Projects/Iperf/, and give it a try.  Basically it will blast tcp packets from your 'server' to the host and record a very accurate measurement. Place one of the computer on the Router and the other behind the hub and see what type of speeds you get.. 

I will try this in lab tomorrow.. I have a Cisco 515e and an older Intel 10baset hub that I could give it a shot..  Here is why I think that this is happening.. Once the buffer fills.. then you see the drop in speed down to, my guess would be, 10Mbit/s or less..  Seeing how the test file is less than the buffer on the WRT54g, i am running DD-WRT, you would see higher speeds because you are pulling to the router faster..  As far as the TMN server knows your router is the device making the request.  So once all of the traffic has been retrieved, the test is over.  I am going to bet if you were to test with a larger test you wouldn't see more than the hub speed. 

[anox195]

I dont think that the hub is allowing those speeds.. my guess is that the router is buffering the IP packets for you there by creating a caching effect..  The best way to see if this is happening would be to run a throughput test on the LAN itself.  Download something like iperf, http://dast.nlanr.net/Projects/Iperf/, and give it a try.  Basically it will blast tcp packets from your 'server' to the host and record a very accurate measurement. Place one of the computer on the Router and the other behind the hub and see what type of speeds you get.. 

Test in-progress.