Anon0
@army.mil
| reply to Nightfall
Re: Zero justification for it
said by Nightfall  :My main point in my post, in case you missed it, was that the total amount of bandwidth used per account and per node, and how much is available is there. The ISPs have been keeping track of this info for years.
But that doesn't really matter. ISPs are making the claim that they have to enforce bandwidth caps on users or their system will be unable to handle the extra load due to demands for things like streaming video. Who cares how much the individual users are using? What matters is how much bandwidth there's still available for everyone. Even if 3% of users are using more bandwidth than everyone else, if there's still more than enough bandwidth for all users then ISPs cannot say their infrastructure will be overloaded. It's twisting the statistics to make them say what you want them to. |
|
jp10558
Premium
join:2005-06-24
Willseyville, NY
| reply to espaeth
That's all great, but none of that would push the pricing the way the carriers are trying to go. Specifically, maintaining the equipment is going to cost the same whether someone uses 5GB a month or 50GB a month. Caps don't address higher line costs.
I mean, suppose everyone decides to live within the caps? That still won't pay for upgrades to get the faster speeds, more users, or even just increased maitenence costs.
--
Opera 9.51(Build 10081); Windows XP Pro SP3;Intel C2Q6600; 3GB DDR2 1066; 1M/128k DSL; Antivir Personal; Comodo Firewall Pro 3;Proxomitron 4.5j Sidki 2008beta,GPG ID:0x0A1C6EE3 |
|
Nightfall
My Goal Is To Deny Yours
Premium,MVM
join:2001-08-03
Grand Rapids, MI
 ·Site5.com
 ·AT&T Midwest
 ·Comcast
| reply to KrK
said by KrK :That's going to change though. More and more "regular users" are going to be moving up the ranks when they start ACTUALLY using their connections for services appearing in infancy on the Net now. The problem is that ISP's have enjoyed and telecom providers have exploited for profit for years the people who buy high speed lines and then use very little. (Hell even I fall into this class.) So now they are scared not so much for network capacity but for profits as they realize that soon the gravy train will be ending and regular Joe Sixpack connections will be starting to use their connections for actual useful services and actually want to use more then 5% of the bandwidth they pay for.... I have to agree with you. The ISPs profits are going to sink even further when this day comes. Once again though, I am not arguing this point at all. Merely that the data these ISPs are keeping on their customers and the "congestion" on their network is well documented and kept track of. |
|
KrK
Heavy Artillery For The Little Guy
Premium
join:2000-01-17
Tulsa, OK
·AT&T Yahoo
·AT&T DSL Service
·Cox HSI
·AT&T Southwest
| reply to Nightfall
That's going to change though.
More and more "regular users" are going to be moving up the ranks when they start ACTUALLY using their connections for services appearing in infancy on the Net now.
The problem is that ISP's have enjoyed and telecom providers have exploited for profit for years the people who buy high speed lines and then use very little. (Hell even I fall into this class.)
So now they are scared not so much for network capacity but for profits as they realize that soon the gravy train will be ending and regular Joe Sixpack connections will be starting to use their connections for actual useful services and actually want to use more then 5% of the bandwidth they pay for....
--
"Regulatory capitalism is when companies invest in lawyers, lobbyists, and politicians, instead of plant, people, and customer service." - former FCC Chairman William Kennard (A real FCC Chairman, unlike the current Corporate Spokesperson in the job!) |
|
espaeth
Digital Plumber
Premium,MVM
join:2001-04-21
Minneapolis, MN
 ·voip.ms
·Vitelity VOIP
·Callcentric
·VoiceStick
·ViaTalk
·Comcast
·Embarq
| reply to wentlanc
said by wentlanc :I don't completely disagree with you either. But the difference between a 4x throughput gain over copper, and the nearly infinite gain through fiber is vastly different. I would rather see cable operators spend a little more, like Verizon is doing, to ensure that the investment will actually last. What's the driver to fund that expansion? Verizon is leveraging their successful wireless division to help finance FiOS, and the gold pot at the end of that rainbow is being able to get into the video business. The cable companies can already deliver video, phone, and Internet over their infrastructure. Not to mention they already have the materials, techs, and training programs in place to handle their copper infrastructure.
said by wentlanc :I'm starting to believe that the cable operators may be looking to protect their investment into DOCSIS v3 with caps. Not quite -- they're limiting how much they need to pour into infrastructure upgrades because the building blocks aren't nearly as cheap as at the head-end. My point earlier is that while ISP to carrier connections have gone from 45mbps DS3s and 155mbps OC3s in the late 90's to 1000mbps and 10,000mbps connections that maybe increased 100%-150% of the costs of the upstream circuits they had a decade ago (for a 1000 - 10000% gain in capacity). In the same timeframe the building blocks for expanding DSL and DOCSIS haven't changed, and thus aren't cheaper to the same degree as head-end bandwidth.
In the last decade we've gone from 600-1000 homes sharing a downstream DOCSIS channel to about 100 homes today. The BPON FiOS deployments are serving 16 homes per distribution, the GPON buildouts are serving 32. If cable companies got down to those numbers for homes per node they can easily match anything FiOS will throw in the near term. With 38mbps per 6MHz channel, there's 4.75gbps of capacity in a 750MHz plant. The problem is there is nothing for Video or Voice services that is driving the need for that kind of expansion, so if HSI is going to be the driver it's going to need to be the source of the funding. |
|
wentlanc
You Can't Fix Dumb..
join:2003-07-30
Maineville, OH
| reply to Dogfather
said by Dogfather :I disagree on the investing part. These providers reinvest a good portion of their revenue. Comcast is going DOCSIS 3, SDV, moving away from analog. Verizon obviously is dropping over $20 billion on fiber. AT&T is moving to 40 gig and then 100 gig. I don't completely disagree with you either. But the difference between a 4x throughput gain over copper, and the nearly infinite gain through fiber is vastly different. I would rather see cable operators spend a little more, like Verizon is doing, to ensure that the investment will actually last. I think that by the time they get DOCSIS v3 in place in most markets, they will be looking to upgrade again.
I'm starting to believe that the cable operators may be looking to protect their investment into DOCSIS v3 with caps. If they planned for these upgrades, but the average internet usage causes them to surpass their planned capacity, they would have to invest again to be able to compete against AT&T and Verizon.
Hmmm.......
cw |
|
Dogfather
Premium
join:2007-12-26
Laguna Hills, CA
·Cox HSI
 ·Verizon FIOS
·Cox VOIP
·ViaTalk
·RoadRunner Cable
·MegaPath
·Verizon west (ex G..
·Time Warner VOIP
| reply to wentlanc
I disagree on the investing part. These providers reinvest a good portion of their revenue. Comcast is going DOCSIS 3, SDV, moving away from analog. Verizon obviously is dropping over $20 billion on fiber. AT&T is moving to 40 gig and then 100 gig.
Moore's law is working out and keeping these providers well ahead of capacity which is why when we do see data, like that of Bell Canada, there is zero evidence of congestion. And while there may be some upstream channel saturation in come neighborhoods in some cable systems, the problem is certainly not catastrophic enough to warrant draconian caps.
This is why I would 100% agree with you that this is a money grab in that this is all about protecting video revenues. AT&T, Time Warner, Comcast, all the big names in "capping" have pay-per-view revenues to product and even minute spent watching video from other provider is a minute not watching it on their own services. |
|
fireflier
Coffee. . .Need Coffee
Premium
join:2001-05-25
Limbo
·Skype
| reply to Dryvlyne
Indeed. If as the comcast network engineer showed, they know down to the node and account who is running flat-out, they should be dealing with those customers rather than using them as justification to impact everyone.
--
Tradition: Just because you've always done it that way doesn't mean it's not incredibly stupid. --despair.com |
|
wentlanc
You Can't Fix Dumb..
join:2003-07-30
Maineville, OH
| reply to espaeth
said by espaeth :They're reaching the top end of their distribution technology. The 38mbps downstream DOCSIS channel is the EXACT SAME 38mbps channel capacity they had when they started rolling this stuff out in 1999. Sure, they've done a vast many node splits over the years to reduce subscriber density per channel, but that doesn't change the fact that when they rolled this out they had all kinds of headroom in contrast to 1.5mbps subscriber connections. DSLAMs as well could be fed off T1s to deliver 512k and 1.0mbps connections back at the early part of the decade. They are at the point now where they are coming to the end of the excess capacity in the plants that they can just dial up. Growth beyond this point is going to require real physical infrastructure upgrades which will require lots of real money. And this is exactly the problem. Instead of actually investing in a future for their network, they planned to try to squeeze the life out of their existing network. And now they want to charge people MORE, for the same network with the same issues, regardless of the fact that the money that they collect CANNOT make the issue go away. If they are imposing fees to more than just those users, then there was no issue in the first place, and this is just a money grab.
cw |
|
Nightfall
My Goal Is To Deny Yours
Premium,MVM
join:2001-08-03
Grand Rapids, MI
·Site5.com
·AT&T Midwest
·Comcast
| reply to RARPSL
said by RARPSL :said by Nightfall :Here in the city of Grand Rapids, the top 3% of their customers use more bandwidth than the bottom 97% combined. So. That is a FUD claim unless you ALSO back it up with the amount the percentage of total available bandwidth that is actually being used. If the total usage is 95% of the available bandwidth the fact that those 3% are using more of the 95% than the other 97% of the users is not important. Only if the usage is 100% of capacity does the split possibly become an issue. I never made a claim that Comcast bandwidth in my area was an issue. So what FUD claim did I make?
Secondly, you don't wait until you are using 95% of the bandwidth in an area before you add more. A well managed network has plenty in reserve, like about 20%, in case something bad happens. They also have multiple points of entry. Thats besides the point though...
My main point in my post, in case you missed it, was that the total amount of bandwidth used per account and per node, and how much is available is there. The ISPs have been keeping track of this info for years. |
|
Dogfather
Premium
join:2007-12-26
Laguna Hills, CA
·Cox HSI
·Verizon FIOS
·Cox VOIP
·ViaTalk
·RoadRunner Cable
·MegaPath
·Verizon west (ex G..
·Time Warner VOIP
2 edits | reply to espaeth
Doesn't mean that it is eroding the profits either. And certainly there is no right to a certain margin and with increases in volume (eg number of total subscribers) we would expect to see margins lower (with overall profit higher), the result of competitive forces.
If the ISPs want to grow some balls and argue succinctly that they're capping to raise profit margins; let them. But they're not. They're crying congestion and "preserving the customer experience" while there is no evidence any such widespread problems exist. And again, when we are shown the data, the opposite turns out to be true; there is no capacity crunch. |
|
espaeth
Digital Plumber
Premium,MVM
join:2001-04-21
Minneapolis, MN
·voip.ms
·Vitelity VOIP
·Callcentric
·VoiceStick
·ViaTalk
·Comcast
·Embarq
| reply to Dogfather
said by Dogfather :There is no evidence of congestion on these networks and when ISPs are forced to turn over data it turns out the exact opposite is true. Congestion isn't the motivation for the caps because there is no congestion. Again, all that means is that ISPs have been pacing their upgrades reasonably well to avoid major congestion. It doesn't mean that their increasing CapEx and OpEx costs from said upgrades aren't eroding the profit margins of the service. |
|
Dogfather
Premium
join:2007-12-26
Laguna Hills, CA
·Cox HSI
·Verizon FIOS
·Cox VOIP
·ViaTalk
·RoadRunner Cable
·MegaPath
·Verizon west (ex G..
·Time Warner VOIP
| reply to espaeth
There is no evidence of congestion on these networks and when ISPs are forced to turn over data it turns out the exact opposite is true. Congestion isn't the motivation for the caps because there is no congestion.
With that we'll just have to agree to disagree. |
|
RARPSL
join:1999-12-08
Suffern, NY
| reply to Nightfall
said by Nightfall :Here in the city of Grand Rapids, the top 3% of their customers use more bandwidth than the bottom 97% combined. So. That is a FUD claim unless you ALSO back it up with the amount the percentage of total available bandwidth that is actually being used. If the total usage is 95% of the available bandwidth the fact that those 3% are using more of the 95% than the other 97% of the users is not important. Only if the usage is 100% of capacity does the split possibly become an issue. |
|
fiberguy
My views are my own.
Premium
join:2005-05-20
| reply to jaminus
said by jaminus :Why slow down high users, when you can just make them pay more? In the DSL model I'd agree with you.. make them pay more. Still, I think it should be close to the Fair and Flexible model that Sprint once had with cell phone use. If you go over, they just upgrade you to the next plan for that month.
As for cable, it's a little different. The lines were built for residential use.. they weren't designed for heavy users. In fact, some cities/systems won't let residential homes sign up for business accounts just because of the implied heavy use in that area. The problem is that you have to cut the heavy users back to some degree, not necessarily allow an overage fee, because in a node model, and say you luck out and have a lot of heavy users in one area, they will slow the node down for everyone. You can split the node, but in some areas that's either a costly move OR not worth it if there is a high churn rate in the area.
In some nodes where there are higher rental homes/apts, you split the node and the people move to another node, that split was for nothing. And, in lack on contracts, you never know if the node split would be necessary for too long.
One thing that COULD work is a contract term on heavy users too. If a node split has to be done, at least those causing high use would be held to that contract, no ETF, and paid to term, for the trade of node splitting.
Yes, I realize that the last option is not user friendly, but neither is capital expense for a few people either.  Some operators are actually okay with seeing a few problematic customers become someone else's problem.
I still, also, say that we're in a turning point in the history of the internet where the service and the content are growing at a fast paced and a bit off balance.. I think all of this issue of heavy use will go away in the next 8 years. |
|
jaminus
join:2004-10-14
Arlington, VA
| reply to TKJunkMail
If I'm interpreting what you're saying correctly, then I think you have it wrong.
Consider your average broadband user's consumption habits. They download large files infrequently, and the main thing they do is load websites, download songs, and the like.
Why would the typical user want a relatively narrow pipe they could saturate, when most people simply aren't downloading constantly?
Overselling hasn't emerged because consumers are stupid. It has succeeded as a business model because it makes sense. Even as a relatively high bandwidth user, I prefer a fast 16mb with monthly usage caps to a slower, uncapped connection. I want to be able to download 7 Gigs in an hour, and I don't have a problem if my ISP oversells its nodes to a point.
Residential broadband providers don't ever claim you can use your connection all the time. BroadbandReports.com users may want an uncapped pipe, but it'd mean much higher prices, which is unacceptable to the typical broadband user. |
|
espaeth
Digital Plumber
Premium,MVM
join:2001-04-21
Minneapolis, MN
·voip.ms
·Vitelity VOIP
·Callcentric
·VoiceStick
·ViaTalk
·Comcast
·Embarq
| reply to Dogfather
said by Dogfather :Internet VoIP control faces 2 hurdles. The first is very simple, the FCC is watching for this like a hawk. So the FCC will watch for voice interference with hawk-like vigilance, but look the other way for anti-competitive video behaviors?
said by Dogfather : Second, cable is crushing indy VOIP with little effort and they're not doing that with video. Are Internet "indy" VoIP subscribers converting to MSO-based services, or are MSOs simply able to reach customers that the Internet companies can't add because they'll send techs out to handle the installation and provide on-going support via phone and truck rolls?
said by Dogfather :As for your broadband price model, you're model requires that the per unit cost of bandwidth doesn't drop which it certainly does, keeping per user bandwidth costs the same as consumption increases. You also have to factor in why and where bandwidth is getting cheaper. Physical connections aren't really getting any cheaper, in fact, physical plant costs continue to rise slightly. Commodity Ethernet has certainly factored into reduced costs on the carrier / infrastructure side. Backbone carriers have largely abandoned expensive circuit-switched interfaces and transitioned to GigE and 10GigE handoffs. We have private fiber between some of our key locations today, it costs us about $5k/mo for the fiber and that includes bundled maintenance (ie, they have trucks to repair cuts). In the last couple years that cost has actually gone up from about $4k to $5k to cover the increased maintenance expenses, but over that same time 10GigE interfaces have come down in price so we have been able to upgrade the capacity of the connection 10x with a single CapEx equipment interface upgrade. My cost per mbps has decreased substantially, but my actual recurring cost of the connection has gone up.
The cost to deliver the connection to homes is the same story. The labor force to maintain the physical plant isn't getting any cheaper, with copper prices climbing the cost of cabling isn't getting any cheaper, gas prices are driving up the cost of operating the maintenance vehicle fleet, and energy costs are making the plant more expensive to operate overall. There is no driver in the DSL or DOCSIS world to drive bandwidth prices down at the same rate as the backbone. The connections are more distributed, there are more devices to upgrade to increase the speed across the path, and the gains than can be achieved are still somewhat minimal. DSL you can go from 7mbps up to 20mbps by pushing out a deeper FTTN infrastructure, and on DOCSIS plants upgrading to v3 gets you the ability to increase capacity 4x by jumping from 38mbps to 152mbps downstream. Not only is it more expensive at the edge to achieve these gains, but the scaling factor isn't even remotely close to the 1,000mbps to 10,000mbps bump that can be achieved with minimal effort at the head-end.
said by Dogfather :We've seen from Bell Canada's data, that even with the increase in traffic, there is no capacity crunch and providers like AT&T who in confidence reiterate that they have no capacity crunch expand from 10 gig to 40 gig and eventually to 100 gig while their data services margins increase (according to their annual report). That's because providers like Bell Canada have been upgrading their infrastructure all along, albeit at a pace slower than the growth of demand. You make these statements like providers should wait for things to hit 100% and then start talking about imposing different billing. You have to realize that at the backbone level the tolerable level of congestion within a network is 0%. Once circuits start to hit 60% capacity carriers start to prep additional capacity so that it can be in place before the 75-80% boundary is crossed. If backbones ever reached the point of any kind of perceivable congestion latency-critical applications like gaming and voice would cease to function at acceptable performance levels.
said by Dogfather :It's certainly no coincidence that after a decade of providing unmetered service that only after rolling out their own video product (U-Verse) are they moving to caps. They're reaching the top end of their distribution technology. The 38mbps downstream DOCSIS channel is the EXACT SAME 38mbps channel capacity they had when they started rolling this stuff out in 1999. Sure, they've done a vast many node splits over the years to reduce subscriber density per channel, but that doesn't change the fact that when they rolled this out they had all kinds of headroom in contrast to 1.5mbps subscriber connections. DSLAMs as well could be fed off T1s to deliver 512k and 1.0mbps connections back at the early part of the decade. They are at the point now where they are coming to the end of the excess capacity in the plants that they can just dial up. Growth beyond this point is going to require real physical infrastructure upgrades which will require lots of real money. |
|
jaminus
join:2004-10-14
Arlington, VA
| reply to fiberguy
Why slow down high users, when you can just make them pay more?
I'd be fine with a base price of $50 per month up to 200GB, then $0.25 GB at normal speed after that -OR- you get slowed down to the lowest priority during peak hours where congestion is an issue.
Make the hogs (like myself) pay and they'll either cut back, or at least they'll bear the burden of the added strain they induce on the last-mile. |
|
Dryvlyne
Far Beyond Driven
Premium
join:2004-08-30
Newark, OH
| reply to Nightfall
All the more reason for those top 3% users to be migrated to a business tier or risk being disconnected. There is no way in hell any ISP can justify usage-based billing to make it "fair" to everyone who uses the Internet. Quite the contrary, it would be quite UNfair for the vast majority of users to have to pay for the abuse of such a small minority. Furthermore, as you've noted, ISP's have the means to identify specific users that are placing such extreme demands on their network.
The idea that it would somehow be 'fair' for the ISPs to just throw everyone into a new pricing model because of the actions of a small minority is simply ridiculous! |
|
fiberguy
My views are my own.
Premium
join:2005-05-20
1 edit | reply to Nightfall
The very same is said here in the Twin Cities.. That statement about the top 3% using about the same, or more, than the rest of the 97% is echoed around in different systems..
If people actually saw these numbers in person, at the actual computer screen showing the data, people would flip and often not believe it.
Personally, I have no issue with caps, but would like to see them higher.. Comcast wants to do a 250gb cap, rather, I think 500gb is more reasonable, maybe a little higher. Further, passing the cap doesn't mean charges or cut off, personally, I'd rather see the connection slowed to that of a 1.5/256 line. If people want to run wide open connections for everyone else to take files from them.. fine. But when their line becomes almost unusable to the account user him/herself, maybe they'll think twice about opening up their connections to everyone and keep their bandwidth to themselves.
Further, as networks continue to be upgraded, the caps should also reflect reality. |
|
