Maybe if vonage wasn't using TFTP to get their configs, they would find themselves blocked less. TFTP is blocked on many networks due to the fact its a popular mechanism for trojans to download and install their ad/spyware.
There are many other ways to make Vonage not work very well without blocking it. If you pay for a best effort internet connection with me (which almost all cable modem service is) you aren't promised anything. You can still connect to vonage, but you have no guarantee of anything. If i set all premium traffic (stuff people are actually paying extra for) to a higher QOS level and there is congestion anywhere, youre VOIP call goes bye bye. I'd like to see the FCC fine that.
Vonage is a parasitic application that rides over the infrastructure of others. They have no agreements to pay anyone for the infrastructure they are taking advantage of. On my network, P2P is more preferred than Vonage so you better stop downloading before trying to make that call.
And then you have the geniuses who subscribe to vonage and then try to sue the cable company because their 911 call failed. You aren't paying me for E911 service. You better take it up with your voice provider, Vonage. If they want me to provide that service for you, they need to pay for it, as I do.
This isn't any kind of victory for VOIP. It's just another sign that IP is going to be regulated.
The CIX (commercial internet exchange) was started in August of 91. It was the first place that ISP's could exchange 'commercial' traffic. The founding members were Cerfnet (at&t now), PSI, and Alternet (uunet).
I know for a fact that Cogent is having issues even getting public peering. So they have to buy transit. Even if they get a record breaking 100 dollars a meg transit deal, they can't make money.
This is going to be interesting. Oh what a tangled web we weave.
Lets look at this pricing model. Unoversubscribed
100meg per customer. OC192 backbone, about 10000 megs. 10GB's. That's alot isn't it? Oh wait..
10000mb/100mb= 100 customers? 100 customers * $1000 a month? 100,000 a month revenue? How were they paying for that OC192? Unless its free, someone lying about something.. hell, lets say they get total free peering in each city. Now they can do 100 customers in each city. Doesn't work. How are they going to pay for it? oversubscription.
As A network architect for a large internet backbone, I can tell you one thing about this.
It looks good on paper. It sounds good in the press. It makes your stock prices go up. But it doesn't move your traffic any quicker. The truth of the matter is, the people who this product is targeted at are the same people who are paying 1500 a month for a T1 that they use 1/10 of for 8 hours a day 5 days a week. Ok.. So now CogentCO is going to go in to major cities and wire all the buildings.. oh.. I dunno.. lets say.. LA for example. In 1 square mile of downtown LA they could possibly pick up.. oh.. I dunno.. lets say 3000 customers, all getting 100 megs of traffic.
Since none of the sites these folks are trying to get to are located on the Cogent network, the traffic is going to have to go somewhere.
The majority is going to go to UUnet, Sprint, Exodus, and Abovenet. How much traffic was that for LA again?? Oh yeah.. 3000 x 100mb.. 300,000 mb. WHAT??!!! C'mon.. I know you can't really push 100mb across 100baseT.. And I know they aren;t going to have 3000 customers there.. but if just one customer users their full pipe, they have 100 meg of traffic destined for other networks. 10 customers? 1000 meg. The exchange points get busy.
Maybe they can haul these thousands of megs of traffic back and forth across their OC192 backbone all day long. Fine. Wait til they try to give that traffic away. Thats where your bottle neck comes in. Theres only one way to stop this and to create the INTERNET of the Future. The one we dream about. Everything on demand. Right at your fingertips. No delay. Speed of light baby.
It has to be 1 network. None of this exchange point stuff. I know this isnt a very popular opinion.. but its true. If we continue to design and build many different networks and tie them together with different technologiesl, we're never going to get it right. Oh.. damn.. can't do that. Justice department won't let us. So live with it. It's going to get a lot worse.
Slashdot Mirror
Maybe if vonage wasn't using TFTP to get their configs, they would find themselves blocked less.
TFTP is blocked on many networks due to the fact its a popular mechanism for trojans to download and install their ad/spyware.
There are many other ways to make Vonage not work very well without blocking it. If you pay for a best effort internet connection with me (which almost all cable modem service is) you aren't promised anything. You can still connect to vonage, but you have no guarantee of anything. If i set all premium traffic (stuff people are actually paying extra for) to a higher QOS level and there is congestion anywhere, youre VOIP call goes bye bye. I'd like to see the FCC fine that.
Vonage is a parasitic application that rides over the infrastructure of others. They have no agreements to pay anyone for the infrastructure they are taking advantage of. On my network, P2P is more preferred than Vonage so you better stop downloading before trying to make that call.
And then you have the geniuses who subscribe to vonage and then try to sue the cable company because their 911 call failed. You aren't paying me for E911 service. You better take it up with your voice provider, Vonage. If they want me to provide that service for you, they need to pay for it, as I do.
This isn't any kind of victory for VOIP. It's just another sign that IP is going to be regulated.
But does nature have a patent on the frog?
The CIX (commercial internet exchange) was started in August of 91. It was the first place that ISP's could exchange 'commercial' traffic. The founding members were Cerfnet (at&t now), PSI, and Alternet (uunet).
This has been a public service announcement.
I know for a fact that Cogent is having issues even getting public peering. So they have to buy transit. Even if they get a record breaking 100 dollars a meg transit deal, they can't make money. This is going to be interesting. Oh what a tangled web we weave. Lets look at this pricing model. Unoversubscribed 100meg per customer. OC192 backbone, about 10000 megs. 10GB's. That's alot isn't it? Oh wait.. 10000mb/100mb= 100 customers? 100 customers * $1000 a month? 100,000 a month revenue? How were they paying for that OC192? Unless its free, someone lying about something.. hell, lets say they get total free peering in each city. Now they can do 100 customers in each city. Doesn't work. How are they going to pay for it? oversubscription.
As A network architect for a large internet backbone, I can tell you one thing about this. It looks good on paper. It sounds good in the press. It makes your stock prices go up. But it doesn't move your traffic any quicker. The truth of the matter is, the people who this product is targeted at are the same people who are paying 1500 a month for a T1 that they use 1/10 of for 8 hours a day 5 days a week. Ok.. So now CogentCO is going to go in to major cities and wire all the buildings.. oh.. I dunno.. lets say.. LA for example. In 1 square mile of downtown LA they could possibly pick up .. oh.. I dunno.. lets say 3000 customers, all getting 100 megs of traffic.
Since none of the sites these folks are trying to get to are located on the Cogent network, the traffic is going to have to go somewhere.
The majority is going to go to UUnet, Sprint, Exodus, and Abovenet. How much traffic was that for LA again?? Oh yeah.. 3000 x 100mb.. 300,000 mb. WHAT??!!! C'mon.. I know you can't really push 100mb across 100baseT.. And I know they aren;t going to have 3000 customers there.. but if just one customer users their full pipe, they have 100 meg of traffic destined for other networks. 10 customers? 1000 meg. The exchange points get busy.
Maybe they can haul these thousands of megs of traffic back and forth across their OC192 backbone all day long. Fine. Wait til they try to give that traffic away. Thats where your bottle neck comes in. Theres only one way to stop this and to create the INTERNET of the Future. The one we dream about. Everything on demand. Right at your fingertips. No delay. Speed of light baby.
It has to be 1 network. None of this exchange point stuff. I know this isnt a very popular opinion.. but its true. If we continue to design and build many different networks and tie them together with different technologiesl, we're never going to get it right. Oh.. damn.. can't do that. Justice department won't let us. So live with it. It's going to get a lot worse.