Re: Average Usage Per Month By Tier
 

They probably anticipated it would have the same proportionate increase as similar upgrades to downstream speeds.

Wrong, obviously, as most upstream usage is still P2P, so will fill whatever pipes they provide. With some exceptions most people don't make attempts to saturate their downstreams 24x7 because they can as they only grab stuff they can use, while serving a popular torrent will happily unattended fill the upstream 24x7.

Re: Average Usage Per Month By Tier
 

That's very much a no **** Sherlock situation. How did VM miss it and isn't the upstream caught by the shaping?

Re: Average Usage Per Month By Tier
 

While "hindsight" allows a far more accurate "load forecasting model" ;) it does imply that VM naively ignored a wealth of "trend analysis" in the public domain over recent years and/or lacked the necessary sophisticated DOCSIS modelling software to derive more accurate predictive load forecasts by area?

Clearly we'll never know the details as to what "wet finger" the VM's Systems Architect employed or financial constraints that presumably prevented engaging the usual typical University based Research Projects or Independent Consultancy available.

Interesting to see what contingency and forward planning has now been effected for the predictable increased demand during the 2012 Olympics.
Again we'll never know whether VM have "learnt from their mistakes" until the "event" occurrs! :(

Re: Average Usage Per Month By Tier
 

I suspect the upgrade was marketing led based on what Infinity offers as upstream. The amount of science would be limited to whether or not the waved finger should be licked.

Re: Average Usage Per Month By Tier
 

I can't help but agree. I also strongly feel that the Systems Architect(s) were very much financially constrained. The alternative proposition is that they are fools, and I find that hard to believe.

Forgive my ignorance, but based on my limited understanding of DOCSIS, I believe that asymmetry is by far its greatest weakness, and the magnitude of asymmetry is only likely to trend upwards with time?

Also, my understanding is that the overall ratio of traffic is trending towards an increase in upstream, as more and more users discover the joys of legitimate file sharing through p2p? (Or perhaps this is counterbalanced by a proportional increase in legitimate downloads from Usenet, along with video streaming)?

In any case, is there not an argument for an increase of the magnitude of traffic shaping for p2p? (Ignoring for the moment all the problems associated with the misidentification of packets.)

Frankly, if it were up to me, I would drop every single Bittorrent packet (that I could recognise).

Phil.

Re: Average Usage Per Month By Tier
 

I have to declare a personal interest here.

My own connection was performing just fine for ages then suddenly there was a marked degradation - sometimes to quite abysmal performance levels.

From looking at my TBB monitor and some speedtests I'm pretty much convinced that a lot of the congestion is down to a single user - the onset and end of periods of abysmal performance are quite sharp so I don't see what would cause co-ordinated usage of many users plus speedtests taken at such times show a very low upstream. I'm pretty much sure that a single user has taken 100Mbps and is hammering torrents.

I've got an outstanding question on the VM community board on the eta of a fix for the fault reference they gave me when I complained about the congestion. If they don't come up with something reasonable (ie this month) I'll be calling in to downgrade as I don't see paying for 50Mbps when a lot of the time they can't deliver 10Mbps. This has completely destroyed any faith I had in VM and I'll almost certainly be moving to Infinity when it becomes available to me. Of course being a BT product that will probably be poor too but I've seen few horror stories yet.

Re: Average Usage Per Month By Tier
 

What is bizzare as well that although docsis is asymmetrical in a big way the allowed usage on the upload vs download before STM kicks in is almost symmetrical 5gig vs 4 gig. On my 30mbit connection I can hit the download limit very quickly, but the upload limit would take many hours of full speed.

Whats more common then realised is streaming videos from home, uploading videos to youtube, hosting gaming servers, hosting web hosting from home (sad but true), then there is generic upload stuff like uploading photos etc. Also the increasing trend of online data storage servers,for things like backups.