It seems like just about any video I watch lately has bufering issues. Plays about 3 or 4 seconds buffers... and stays stuck. I can pause it and it will VERY SLOWLY load the video. Doesnt matter what type of video I.E Music or whatever, everything. I have a 20mb/s FIOS connection and I should see no problem even with 1080p videos.

HULU for instance plays fine along with other networks like grooveshark, but you tube since being bought by Google seems to just plain out suck iht.



Anyone ?

 

Is that round trip time to a location < 50 miles away correct ?

With 236 ms of delay you have delay on par with a VSAT link, just locally. Add the additional time to get to a youtube hosting site, and it only gets worse.

If that time is correct, that is why large data transfers are running slow.

 

That's almost 2000 miles away, still the same pratically.

Its only youtube who has horrible buffering problems. Like I said hulul for instance works fine.

 

ive noticed some vids loading ridic slow lately too. annoying when i want to listen to some songs on youtube instead of downloading them.

 

I think youtube is growing faster than they can update it.

 

Anytime I watch my band the video and the audio are not on the "Same Page"

...zap!

 

This tells me the delay is with VZ, or the peering they have to the internet or your network,

I just did a wireshark this AM to verify youtube leaves the video inside the TCP session, I cannot say what hulul uses ( could be UDP, which does not care as much about delay ).

Edit : Just got a chance to calculate the T-put of your line.

At 190 ms, your max through put for a TCP application with 190 ms of delay is 3 Mbps per flow. So 1 youtube is 1 TCP flow. Does not matter how fast your actual line can run ( could be a 1GB local link ), with that delay standard windows TCP session cannot go beyond 3Mbps ( without any loss happening, which is not the case on the internet ). This means you can have ~ 6 PCs all pulling youtube videos and it would run the same ( or 6 windows on 1 PC, if CPU usage is not an issue ).

Testing my DSL line ( over a noisy WiFi network with the adjacent networks ) has 4.7 Mbps download, and a delay of 48 ms to the chicago server, and with that if I had the local loop speed I could get to 11 Mbps per flow ( again with 0 loss ) , so I can saturate my DSL line with 1 youtube video.


Go figure I can download download data faster than you can, even with a high noise floor on WiFi.

 

First you did less than a 50 mile run. Second the MILLISECOND delay is the delay before any downloading begins, this is why it still shows 20mb, because it is. It does not download while it handshakes, authenticates, or figures response time. It simple is seeing the response time of the server to my computer to do the above actions. I will still push 20mb or 8/20 which would give me a max of 2.5mb/s that I can download at.

What you are trying to refer to are RWIN values MTU,MSS, and TTL.

That little delay is not limiting me to 3mb/s a second as I couldnt download at that if I wanted to. I can however download at 2.5mb/s depending on the server. It always depends on the server and we all know youtube is bombarded.

Response time is done by each particular server, not just your computer. (I run a server at my house, the server response is to see if it's active connection and some auth, handshakes, etc) If we all tested at the same site, we would most likely all have around the same responses times depending on distances, thus assuming we all lived near one another.

I show several servers giving me in the 40ms response times and 20mb/s.

http://www.youtube.com/watch?v=MvY5qMWzRGM

Click the link to it 720p it and make it max screen to see it clearly..