*ALL* ntl Cap Discussion Here Please. (Part 2)

[Nemesis]

A Fresh New Thread for the CAP Discussions

Continued from here

Play nice now ...

[Ignition]

Originally Posted by obvious

Well apparently ntl have blocked programs like DocsDiag from seeing just how under utilised their network is. ntl no longer permit SNMP access to the ISP-side addresses of cable modems, so the -traffic facility of DocsDiag will no longer work.

That program never actually worked properly and any figures it gave were a guesstimate and usually quite on the low side.

Also do you actually know at which point load on a downstream or upstream becomes service affecting? It's not 100%

Just why should we allow customers to observe the traffic on the equipment of ours they connect to. If/when this stuff starts getting enforced, people will be able to check their own traffic. Load on our network isn't really any of their business - we generally require some free capacity for burst, load changes a lot during the day / night, etc.

Impractical, not to mention dangerous. The idea of BT selling to postcodes on the grounds that the CMTS card covering that area is 65% utilised isn't really a big commercial goer.

[PC_Arcade]

I will be affected by the cap, I use Emule and download (copyrighted) films etc. Whilst my PC is on pretty much constantly I probably don't go over the cap by all that much anymore (I used to) as I'm not maxing out my connection 24x7, in fact I hardly ever max out my connection anymore, the last time I did was shortly after upgrading when I downloaded a movie from usenet to see how long it takes at full pelt (~35 minutes!).

NTL are within their rights to introduce and enforce a cap, it's their network and it is up to them to decide, rightly or wrongly, what people can and can't do with it.

I'm against the cap and IF contacted will leave because it's MY choice to do so, yes it's a PITA for me, but as a customer I have the choice and the right to take my business elsewhere if the company I'm currently using no longer meets my needs.

The "cap" argument is a little pointless IMHO, those for the cap, will remain so despite any arguments made from the other side and those anti it will either leave when (if?) it's enforced or learn to live within it.

NTL are a private company, they exist to make money and to keep their shareholders happy. If they lose a couple of thousand (unlikely but possible) due to enforcement of their rules but maintain or increase profitability then their shareholders will be happy and they will have achieved their goal.
What their customers think of them is completely irrelevant, as long as they pony up the cash each month, the ONLY way NTL (or any private company) will re-think a policy is if it hurts them financially.

[th'engineer]

Originally Posted by Ignition

If/when this stuff starts getting enforced,

Do you know something "If "this stuff starts getting enforced or is it me reading too much into the comment

[jtwn]

For the record, docsdiag -traffic definately doesn't work anymore, tried it yesterday after a year when it did work.

[obvious]

Originally Posted by Ignition

That program never actually worked properly and any figures it gave were a guesstimate and usually quite on the low side.

Also do you actually know at which point load on a downstream or upstream becomes service affecting? It's not 100%

Just why should we allow customers to observe the traffic on the equipment of ours they connect to. If/when this stuff starts getting enforced, people will be able to check their own traffic. Load on our network isn't really any of their business - we generally require some free capacity for burst, load changes a lot during the day / night, etc.

Impractical, not to mention dangerous. The idea of BT selling to postcodes on the grounds that the CMTS card covering that area is 65% utilised isn't really a big commercial goer.

As long as ntl keep to the spirit of their promise that they'll only contact customers who are affecting the quality of service of other customers then there's generally not much to worry about.

However, as network load is the lynchpin behind ntl's capping policy then it would be reasonable for a customer to request some evidence of network load if they are ever contacted. This is where I think it could become a customer's business.

Exactly where load becomes service affecting is an interesting point and it's a question that I'd certainly want to know the answer to if I were being accused of significantly reducing other peoples service level. Perhaps you could educate me? (It wouldn't be 65% would it?)

Perhaps the information is commercially sensitive but penetration/homes passed figures are publically available anyway and I think these figures are of more interest to the likes of BT.

[Mauldor]

Maybe it would be worth going over a few Variables to see why people are so heated...

How much Bandwidth does a customer end box have, that is the one serving these XXX users - both download and Upload. Provided with figures, Im sure people can de-duce themself why problems occur when too many people at once download/upload at once.

People can buy a switch/network cards at very cheap prices, they see a Net connection as a extention to this, after all its shoving data around in the same fashion is it not. Maybe CM or ADSL is now the Wrong technology for the home market?? How does China, Sweden or whoever has 10/100mbit connections wire to the home?

Important one this:- It has been said many many times that a person using there connection too much costs the company money - why and how is this?? Ive asked this loads of times and not once had any answer at all. Is it costing money because they can not shove as many people on that UBR??

Warez - Been pointed out many times as the cause, why then do NTL provide a news Server that has Illegal Software on? A Average user can and will find P2P apps and before they know it - also one of them identified as a heavy user - this I have witnessed myself.

Somebody using there connection too much is told about Contention, yet surely on the same breath a person with slow speed, lag in game etc can also be chucked that same bone - thats told "sorry, its a shared service, we never said it would be 3mbit all the time m'kay"....

A cap will NOT solve the problem of the above - a better idea surely is a simple one, basically what it is now, guidelines for useage and somebody abusing this by 24/7 downloading can be warned or dropped - sounds simple to me..

[obvious]

http://www.chetnet.co.uk/faq.htm#uBR is a good starting place

Example

1. Out of the given pool of subscribers, 40% are logged on during busy hour.

2. Out of the 40% subscribers that are logged-on during busy hour, only 25% might be downloading data simultaneously and contributing to the peak activity.

Thus, peak data demand during busy hour is 10% (.4 *.25) of the subscriber base.

Let us assume that a service provider wishes to limit the worst case data throughput per user at peak busy hour to not less than 256 Kbps. Thus it means that for a given line card with a single one 27 Mbps usable 64QAMdownstream channel bandwidth, the total number of simultaneously active subscribers has to be limited to 27000000/256000 ~= 100.

Since simultaneously active subscribers are assumed to be 10% of total subscriber base, we end up with a number of around 1000 subscribers per line card.

I know this doesn't answer all your questions but it's a start.

[rugrat]

I was just typing a post moaning about NTL speeds today.

Thinking it could not get worse when my computer crashed.

Thanks to NTL .any way after a restart it all seems to be working at full

Speed again. So I will have my say about the speed

Upgrate. insteat and the cap who here would pay for the NTL

TV serves if they were tolled they could only watch it so many

Hours a day so why accept what they want to do with an internet

Cap I for one will not if they get in touch with me and say I am

Downloading too much I will tell them to get stuffed cancel

My NTL phone NTL TV and NTL broadband witch in all comes

To well over £100 a month. And get a new isp that as no cap and

Join sky. I think they are looking at the capping from the

Wrong end witch is what you might except from them they say

They will monitor people with a 1mp a day cap for excessive

Use. So take me I download a film its 1.50g when I get my

Speed upgrade that should take me 3 or 4 hours and then my modem

Will not be in use but I will have gone over there 1g a day limit

But my modem as only been on for 4 hours and then we get the people

On a 3g a month limit who are not being monitor say they want to get

The same film to get it there modem will be on 24\7 because of there

Slow speed so who do you think NTL should be monitoring yes NTL

Have got it back to front again.

[obvious]

Originally Posted by Mauldor

.
.

Important one this:- It has been said many many times that a person using there connection too much costs the company money - why and how is this?? Ive asked this loads of times and not once had any answer at all. Is it costing money because they can not shove as many people on that UBR??

Warez - Been pointed out many times as the cause, why then do NTL provide a news Server that has Illegal Software on? A Average user can and will find P2P apps and before they know it - also one of them identified as a heavy user - this I have witnessed myself.
.
.

It would take a lot of inside information to give a full answer but some of ntl's fixed and variable costs can be seen in the quarterly/yearly reports. They are getting around £30Mill ion a month in broadband subscriptions and as I mentioned a couple of times earlier each GB of external bandwidth costs them around 10p.

Bandwidth costs would seem to me to be a small percentage of broadband revenue.

Historically, the provision of newsgroup access within the ntl network saved them money because they wouldn't have to pay those external transit bandwidth costs over and over again for newsgroups. Recently this has been made a mockery of as they now only carry a small fraction of the daily feeds so people end up subscribing to pay news servers outside of ntl's network and not only do ntl end up paying each time the same article is downloaded by a different customer (I dont think they cache nntp to any degree) but customers will tend to download more as the pay providers have a much 'better' selection.

[ian@huth]

Originally Posted by obvious

http://www.chetnet.co.uk/faq.htm#uBR is a good starting place

I know this doesn't answer all your questions but it's a start.

Those figures assume that everyone is behaving like the model which is rarely the case.

If you take that example a little further and try to give users the full speed they are on then the total number of simultaneously active subscribers has to be limited to 27000000/3072000 ~= 9. Any more than 9 and those 9 will be affecting the service of every user on that card. See how easy it is for only 9 users to affect around a thousand other users.
__________________

Originally Posted by obvious

They are getting around £30Mill ion a month in broadband subscriptions

post mostly stripped.

Where does that £30 million a month come from? It seems rather high.

[obvious]

Originally Posted by ian@huth

Those figures assume that everyone is behaving like the model which is rarely the case.

That's a bit of an insult to the model makers who are in the business. It's not about 'everyone' behaving like the model. It's a business of averages. But I guess you're right and they're wrong as usual.

If you take that example a little further and try to give users the full speed they are on then the total number of simultaneously active subscribers has to be limited to 27000000/3072000 ~= 9. Any more than 9 and those 9 will be affecting the service of every user on that card. See how easy it is for only 9 users to affect around a thousand other users.

Your figures do not take into account any statistical probability.

Where does that £30 million a month come from? It seems rather high.

£25 * 1,200,000. It's not an exact figure just a working estimate. If we assume ARPU on broadband to be £20 then it would be £24Mill ion a month.

Originally Posted by ian@huth

Where does that £30 million a month come from? It seems rather high.

As I recall they have about 1.1 million BB customers, of which about 60% are on the lowest tier, 30% on the middle and 10% on the high (actually I think it may have been < 10% on the high).

Working on those figures they are getting about £22 million per month - which is then about £19 million after VAT is removed.

[Stuart]

And bear in mind that the bulk of their costs are likely to be the purchase, installation and operation of the hardware/software used to provide that bandwidth that costs them £0.10 a gig to the user. After all, there are a hell of a lot of staff, and a hell of a lot of equipment between the user and NTLs link to the net. I have it on fairly good authority that it can cost NTL thousands of pounds to upgrade the network sufficiently so that that one heavy downloader does not impact all the other users in their area. That one heavy downloader will only pay 37.99 a month. How long do you think they should wait for a return on the investment?

[obvious]

It's impossible for 1 heavy user to affect anyone elses service. It's when they are all doing it at the same time that the model fails and problems occur.

The most recent reports suggest a userbase of 1.2million at least.

60% (720,000) @ £18 = £12,960 ,000
30% (360,000) @ £25 = £9,000, 000
10% (120,000) @ £38 = £4,560, 000

Revised total would then be £26,520 ,000 a month

I've no idea offhand of ntl's tax situation so will ignore that for the moment.

[ian@huth]

Originally Posted by obvious

That's a bit of an insult to the model makers who are in the business. It's not about 'everyone' behaving like the model. It's a business of averages. But I guess you're right and they're wrong as usual.

The problem with averages is that rarely anyone is average. Take the average man, he has less than two arms, less than two legs, less than two eyes, etc.

Originally Posted by obvious

ThatYour figures do not take into account any statistical probability.

Statistical probability has nothing at all to do with it. That calculation shows that 9 x 3Mb users, 18 x 2Mb users, 27 x 1Mb users or any combination of these whos speeds add up to 27Mb will saturate the 27Mb downstream that you quoted in that example if they all downloaded at full speed at the same time. On a card that has 1,000 users it is very easy to get such a small number downloading at the same time. Just look at the posters in this thread who seem to think that every user on the middle and top tier uses their connection to the max all the time.

Originally Posted by obvious

That £25 * 1,200,000. It's not an exact figure just a working estimate. If we assume ARPU on broadband to be £20 then it would be £24Mill ion a month.

I would imagine that the bulk of NTLs broadband customers are on the lowest tier with a fair number enjoying discounts from the normal prices.

[obvious]

You know Ian that just because the average number of kids per household might be 2.4 it doesn't mean that they're claiming there are actually 2.4 kids in any given household. : DOH :

[ian@huth]

Originally Posted by obvious

It's impossible for 1 heavy user to affect anyone elses service. It's when they are all doing it at the same time that the model fails and problems occur.

You must be using a wierd type of logic to arrive at that conclusion.

If you have a UBR at peak time where the combined usage of customers is greater than the available downstream then just adding that one extra user will lower the speed that others are getting. That is a simple and undeniable fact.
__________________

Originally Posted by obvious

You know Ian that just because the average number of kids per household might be 2.4 it doesn't mean that they're claiming there are actually 2.4 kids in any given household. : DOH :

Which is exactly my point.

[obvious]

Originally Posted by ian@huth

You must be using a wierd type of logic to arrive at that conclusion.

If you have a UBR at peak time where the combined usage of customers is greater than the available downstream then just adding that one extra user will lower the speed that others are getting. That is a simple and undeniable fact.

OMG and you're saying that all the other users have no effect and it's only the 1 user who has an effect. LOL that's pretty funny.

Which is exactly my point

No it was not your point. You've just done a complete U-Turn. You really dont understand averages do you?

[ian@huth]

Originally Posted by obvious

OMG and you're saying that all the other users have no effect and it's only the 1 user who has an effect. LOL that's pretty funny.

That's not what I said and you know it.