1 edit | reply to Karl Bode
Re: Karl, you have any guess The interesting other side of this debate is the fact that content has already all got together and consolidated into just a few players.
•Akamai
•Google
•LimeLight Networks
The majority of all world wide Internet content is delivered through a minority of content distribution players. Far fewer than their are ISPs in the world. These 3 have collectively pushed the network costs away from themselves through strong arm peering ultimatums and ISP bandwidth arbitrage negotiations.
This allows them to push unlimited traffic at little to no network costs for them. While content gains revenue as traffic increases, the flat fee model on the other side of the delivery has not changed and the content paying ISPs for their 1/2 of the traffic is moving to zero. Peering ratios use to keep this balance, but given the limited massive content players and that collective muscle, it is changing.
Note: this is a pretty significant industry change. There use to be two sides paying for "the Internet": Content paid one side and the users paid the other. A very few well organized strategies have moved all the network cost burden to the ISPs and then the users. |
|
disc
join:2005-12-31
Raleigh, NC
1 edit | said by yt:A very few well organized strategies have moved all the network cost burden to the ISPs and then the users. Have ISP operating costs per sub gone up?
I seem to recall that monthly operating costs for cablecos/telcos for high-speed internet is roughly $8/sub. [Edit: And that's a ceiling, some can do it for less.] |
|
 NOVA_GuyObamaCare Kills AmericansPremium
join:2002-03-05 | Makes me feel even happier now to know that Comcast is making money hand over fist when they rape me for $60 every month for allegedly "high speed" Internet. |
|
| reply to yt
The End-User has now changed their habits. An average user used to use around 11GB/month by downloading music, photos, maybe a bit of YouTube. Now the Users are using relatively new applications to stream DVD-Quality video, increasing their monthly usage exponentially. Three standard definition movies a week may use 20GB+...... |
|
 SLDPremium
join:2002-04-17
San Francisco, CA | reply to yt
You do realize that the side you seem to be sympathizing with here, has been overcharging for said bandwidth for years now?!? Boo hoo!!! |
|
|
|
| reply to disc
said by disc:]Have ISP operating costs per sub gone up? For the ultra-high end terabyte downloaders that this is marketed at... I would have to guess a yes on that one. But that was not my point. |
|
 BF69Premium
join:2004-07-28
Camden, TN | said by yt:said by disc:]Have ISP operating costs per sub gone up? For the ultra-high end terabyte downloaders that this is marketed at... I would have to guess a yes on that one. But that was not my point. And a hard cap can put a stop to that. Anyways let's have fun with math $8 per sub, then 7 cents per GB. So a $60 monthly charge for internet would mean someone would have to use 750 GB per month before the ISP loses money. Since Comcast has a 250 GB cap what exactly is the issue? |
|
Romney2012Defeat Obama 2012-Chg we can believe inPremium
join:2002-03-03
USA
kudos:4 | reply to yt
said by yt:The interesting other side of this debate is the fact that content has already all got together and consolidated into just a few players. •Akamai •Google •LimeLight Networks The majority of all world wide Internet content is delivered through a minority of content distribution players. Far fewer than their are ISPs in the world. These 3 have collectively pushed the network costs away from themselves through strong arm peering ultimatums and ISP bandwidth arbitrage negotiations. This allows them to push unlimited traffic at little to no network costs for them. While content gains revenue as traffic increases, the flat fee model on the other side of the delivery has not changed and the content paying ISPs for their 1/2 of the traffic is moving to zero. Peering ratios use to keep this balance, but given the limited massive content players and that collective muscle, it is changing. Note: this is a pretty significant industry change. There use to be two sides paying for "the Internet": Content paid one side and the users paid the other. A very few well organized strategies have moved all the network cost burden to the ISPs and then the users. All good points for those who refuse to believe the ISPs aren't the villains. The content providers are the ones generating tons of traffic while paying next to nothing for it. It is cost shifting pure and simple.
--
My BLOG .. .. Internet News .. .. My Web Page
|
|
Reviews:
 ·DSL EXTREME
| reply to disc
said by disc:said by yt:A very few well organized strategies have moved all the network cost burden to the ISPs and then the users. Have ISP operating costs per sub gone up? I seem to recall that monthly operating costs for cablecos/telcos for high-speed internet is roughly $8/sub. [Edit: And that's a ceiling, some can do it for less.] $8?? I pay Verizon $21.99 for 1M/384K DSL service every month, I wish mine was $8 a month as I'd love to use the extra $13.99 elsewhere.
--
»www.realtor.com/realestateandhom···12798476
»www.exitstrategyrealty.com/custom4.shtml
look for 37130 Calico Blvd |
|
| reply to yt
Did you already drink the Cool-Aid? What makes you think this is actually targeted at ultra-high end terabyte downloaders? |
|
| reply to Romney2012
said by Romney2012:said by yt:The interesting other side of this debate is the fact that content has already all got together and consolidated into just a few players. •Akamai •Google •LimeLight Networks The majority of all world wide Internet content is delivered through a minority of content distribution players. Far fewer than their are ISPs in the world. These 3 have collectively pushed the network costs away from themselves through strong arm peering ultimatums and ISP bandwidth arbitrage negotiations. This allows them to push unlimited traffic at little to no network costs for them. While content gains revenue as traffic increases, the flat fee model on the other side of the delivery has not changed and the content paying ISPs for their 1/2 of the traffic is moving to zero. Peering ratios use to keep this balance, but given the limited massive content players and that collective muscle, it is changing. Note: this is a pretty significant industry change. There use to be two sides paying for "the Internet": Content paid one side and the users paid the other. A very few well organized strategies have moved all the network cost burden to the ISPs and then the users. All good points for those who refuse to believe the ISPs aren't the villains. The content providers are the ones generating tons of traffic while paying next to nothing for it. It is cost shifting pure and simple. You're not mentioning the fact that content from Akamai, Google, or Limelight is cheaper for the ISP as well.
Google and Limelight peer with most ISP which cuts out transit costs for both the content provider (google and Limelight) and the ISP. The Alternative is your ISP paying a transit provider (such as level3 or Cogent) to deliver them traffic while at the same time Google and Limelight would be paying those same transit providers to take their traffic. The direct peering takes out an unnecessary middle man and saves both parties money.
With Akamai they actually pay many ISP's to co-locate their servers within the ISP's network. I can't speak for all ISP's but know that for many they are actually making a profit off of the Akamai co-location. Of course just like with the peering with Google and Limelight the ISP does not have to pay a 3rd party transit provider for upstream access. They also avoid port fees at the exchange and other trivial costs.
Now I can see your point about the content providers pushing costs on the ISP's because it's the ISP's last mile network that is really being pushed to it's limits here and the content providers aren't burdened with the last mile cost. However there is nothing outside of the traffic shaping and blocking of content that the ISP's can do about the growing amount of internet traffic. So that leaves them in two spots. Pay a transit provider for the traffic or save money by bypassing the transit provider and take the traffic for free. Either way they are getting the traffic they might as well choose the cheaper option which is peering with Google, Limelight, and others.
Oh, I should also mention that the peering with Google and Limelight along with the co-location of Akamai servers also lowers the demand for traffic on what people would call the "backbones" which keeps costs for everyone down (well except the last mile ISP LOL.) At the same time though it improves reliability, performance, and latency for the end users on the last mile network (you sitting at home reading this.)
Peering is the future of the internet. There will always been a need for transit but that need is diminishing at a rapid rate. I often hear people talk about "tier 1" backbones but that is an outdated concept and their importance is dropping faster than anyone outside of the industry would believe. |
|
| reply to NOVA_Guy
said by NOVA_Guy:Makes me feel even happier now to know that Comcast is making money hand over fist when they rape me for $60 every month for allegedly "high speed" Internet. Alleged "high speed"? What does that mean? You don't get it? Or is it not what you define as fast? Are you one of those who want 10g up and down for $20 a month? |
|
1 edit | reply to noname10
said by noname10:You're not mentioning the fact that content from Akamai, Google, or Limelight is cheaper for the ISP as well. Google and Limelight peer with most ISP which cuts out transit costs for both the content provider (google and Limelight) and the ISP. ....The direct peering takes out an unnecessary middle man and saves both parties money. This is a myth used to (temporarily) justify this model.
Transit costs for broadband ISPs are a fraction of end to end bandwidth costs. Enabling unlimited bandwidth is never cheaper for the ISP as it does not lower the costs of metro networks or last mile. "Free" peering should only be done with entities that share equal costs of carrying the traffic end to end (1 meter vs 1000KM). If one party sends more than the other, there should be settlement regardless of their Tiered status.
The reason the big 3 have peering with ISPs today is due to their ability to cause great pain to non ISPs by manipulating traffic locations or arbitrage which increase the ISPs cost forcing the issue. There is a cost to deliver bits and the big 3 don't want any part of it. They want those costs pushed to the ISP and in turn the users.
It is easy to blame ones infrastructure provider (TV or data). What most fail to realize is the reality of companies like ESPN raising their prices, or the big 3 no longer paying for their 1/2 of end to end Internet delivery.
While it is easy to point a finger (I guess as I am doing), think through the entire picture vs. the easy "pile on" approach. The Internet economics ARE changing and concerns about the manipulation of this should focus on all involved. |
|
| said by yt:said by noname10:You're not mentioning the fact that content from Akamai, Google, or Limelight is cheaper for the ISP as well. Google and Limelight peer with most ISP which cuts out transit costs for both the content provider (google and Limelight) and the ISP. ....The direct peering takes out an unnecessary middle man and saves both parties money. This is a myth used to (temporarily) justify this model. No you are wrong. It is not a Myth it is a fact. Your entire argument is a fallacy.
Think about this logic here. I say peering saves the ISP's money, and you say that is a myth based on the savings being small? In fact your argument against me supported my position not yours. |
|
| reply to yt
said by yt:said by noname10:You're not mentioning the fact that content from Akamai, Google, or Limelight is cheaper for the ISP as well. Google and Limelight peer with most ISP which cuts out transit costs for both the content provider (google and Limelight) and the ISP. ....The direct peering takes out an unnecessary middle man and saves both parties money. "Free" peering should only be done with entities that share equal costs of carrying the traffic end to end (1 meter vs 1000KM). If one party sends more than the other, there should be settlement regardless of their Tiered status. Ok, I wanted to adress this statement too because it is so asinine. Actually I am having trouble responding to it. You want both the ISP and content provider to to pay MORE MONEYto add a unnecessary middleman network and provide a less reliable and lower quality of service with higher latency at the same time? Are you a child or a union worker?
It actually wouldn't surprise me if your a union telco worker. Please tell me you are for my own amusement LOL. |
|
4 edits | reply to noname10
said by noname10:Think about this logic here. I say peering saves the ISP's money, and you say that is a myth based on the savings being small? In fact your argument against me supported my position not yours. Let me try and clarify it.
• The small (temporary) savings are for some ISPs for the existing traffic.
• Or.. the traffic may already be traversing a peer of that ISP and while the CDN is, and should, be paying for use of a network, it doesn't save the ISP any transit costs. Peers have a balance of trade where either party carries relatively equal load/equal distance of the other parties traffic. (hence the peering relationship vs transit)
• There is a end to end cost of traffic that has first mile, core, metro and last mile. The most expensive part of the network is as it gets close to the user. The metro and access.
• While a CDN may (and I stress may) cut out some national costs, allowing UNLIMITED FREE traffic without any network growth recover costs takes the existing (temporary) savings of not paying transit and 10x, 100x the amount of content that can be sent at no incremental cost.
This in turn takes that temporary transit savings (small piece of the $$ pie) and enables unfunded exponential growth requiring large capital outlay on the metro and last mile.
Most all content is consolidating behind 3 players and companies like the one you work for are pushing for no more network costs (for them). This dismantles existing Internet economics and pushes all network costs to the consumers.
Pure and simple: Free peering should be with peers that have an equal network cost burden of carrying traffic. 10 meters vs 1000 KM is not equal network cost burden. Consumers should not be the only ones paying for the growth of Internet traffic. |
|
4 edits | reply to noname10
said by noname10:Ok, I wanted to adress this statement too because it is so asinine. Actually I am having trouble responding to it. You want both the ISP and content provider to to pay MORE MONEYto add a unnecessary middleman network and provide a less reliable and lower quality of service with higher latency at the same time? Are you a child or a union worker? It actually wouldn't surprise me if your a union telco worker. Please tell me you are for my own amusement LOL. As stated before, the intermediate providers are not the main cost of Internet infrastructure. In fact, most of them provide very little of the network as most broadband networks have backbones these days. Most of the middle transit providers carry the traffic for 1 router hop and are "selling" the broadband networks (very similar to Cogent issues in the past). The middle providers gain the profits without having any of the real costs. This is why the Tier1 peering system is in such a mess these days.
The broadband networks still need to carry this traffic end to end and any minor savings (if any) is temporary and negatively compounded by the impact of an increase growth trend (do to unfunded free access) on the edge.
I am not advocating content pay everyone... I am saying that content should pay someone for network use and balance of trade/proper peering will maintain the economics. Otherwise all costs push to the consumer. |
|
| reply to Winterman
Re: Karl, you have any guess We sell 512Kbit through 14Mbit to our customers. Our highest 3% of customers use about 19GBdown/3GBUp monthly (this one is on 10Mbit). 45% of ALL of our customers use under 3GBdown/month. This is taken from a pool of 239 people (my ISP). I have only been in business for 1 year; excluding the 6 month's prior of surveying customer interest and finding resource. I am sure if I asked the cable or telco they would have roughly the same numbers. Our demographic is college students and typical households. We have no caps or anything. We don't even QoS any services (other than prioritizing http and https for browsing).
I think alot of the telco's and media conglomerate's are trying to flex those numbers a bit to scare everyone into a system as such. I saw one report a while back where the only customers they were getting their numbers from were user's that were male/single/aged 19-28.
BTW our 512kbit package is only $13.49 and we still clear alright. Our typical cost per sub is only $5.50/month. |
|
