Another point to make is that at 8 megabit/s (not that uncommon speed for HSPA), spending those UKP240,000 takes ~11 hours. 40GB of data is approximately what you can fit on a standard Bluray disk.
The problem with ECN is that no major core router platforms support it, neither does the smaller routers. The only "core" networking device I know of that supports ECN is the CPU type routers (7200 et al) that Cisco makes. Their 12000 and CRS-1 do not.
There is little reason to have ECN turned on at end systems, if the intermediate devices that move/buffer packets doesn't set/use the ECN flag.
I agree. Flow routing with 4M flows didn't work in devices engineered in the late 1990ies, and they won't work now. Any script kiddie can create new "flows" by sending random port/dst IP packets thru the device, and it'll fall over and die just like the devices 10 years back did.
We stopped doing flow routing for a reason, it didn't work. Routers need to care about IP addresses and perhaps take into account port numbers to do load sharing between equal cost links, but nothing else. Looking into flows does NOT scale.
In Sweden, typical monthly peak average is between 20 and 400 kilobit/s depending on type of access and type of users.
This is equivalent to around a few GB per month (remember, it was peak bw usage during the day in 5 minute interval) to 40-60 GB per month.
This means that in most markets, 250GB is hit by a few percent of the users, but on the other hand 250GB per month is 0.7 megabit/s average usage, and with 10/10 megabit connections, a user can theoretically hit 10 times that cap, thus I understand why Comcast wants to do this.
Well, older cars are not protected of course (so there's your 50% right there) plus that an immobilizer doesn't protect against someone breaking into your car to steal stuff. It only means it's much harder to steal the actual car.
It became a requirement in all of EU in 1998 that all new cars must have immobilizer fitted from the factory to be allowed to be sold. Just weird that it's not a requirement in north america.
If you're running four full-bgp VPN-customers in your core routers that can handle 2M routes, you've done a major design mistake and that's not the fault of the hardware manufacturer.
Regarding routing table growth, hopefully IPv6 might stifle that a bit as we're going to be running out of IPv4 space in the next 3-5 years and IPv6 space is allocated in much larger blocks requiring fewer routes.
What some might have missed would be the opportunity to sell gfx cards due to all of a sudden being able to do H.264/VC-1 offload/acceleration in linux for HTPC usage.
The first manufacturer to offer this would get a lot of sales due to people using their cards for HTPC. I know I would, because currently I can't play 1080p VC-1 and barely h.264 properly, even on a Core2Duo.
Net Neutrality is not the answer to the problems seen in the US. The correct answer is to make the largest players rent out their infrastructure with bitstream access and LLUB (Local Loop UnBundled).
As soon as other companies can buy access to the customer and sell them services, then the largest players can't offer degraded or bad service, because the customer can go elsewhere. The problem that Net Neutrality tries to solve is a problem because the customers in a lot of areas don't have many companies to choose from. Solve that problem instead of trying to enforce Net Neutrality and the US will be much better off.
Sweden has had a system like this since 1960 or so. Whenever I do anything I give them my social security number (equivalent) and that's all I need. You cannot do anything in Sweden without one, unless you're all cash.
If I pay with a credit card I have to show ID. Identity theft is extremly rare here, I never even consider it. Very little personal business involving identity is done without showing for instance a drivers license (which also contains the Personal Identity Number).
The UK system of showing a gas bill or alike is just weird for someone like me.
Ok, one doesn't build residential broadband like ones builds T1 circuits to business. Everything you know about that you can just scrap when it comes to residential capacity.
Residential broadband over fiber is built using (ten)gigabit ethernet over dark fiber or CWDM at the distribution, and using 10/100 ethernet over fiber for the access fiber. This is the only way to make it low-cost enough for households to be interested.
Sweden is a country the size of californa with 9 million people. It is estimated that the cost of supplying fiber to within 30 meters of each household in Sweden costs approx $10bn, and if you skip the last 10% of households, you halve that cost to $5bn.
This is just dark fiber, not anyone actually lighting it up. But then again, comparing it to what roads cost per year, it's not very much money.
Here in Sweden we have a lot of municipality owned power companies that have entered the broadband business, some supplying fiber to the curb, some going all the way and connecting private houses and offering services.
Generally, these entities have little clue on what is important when offering services, whereas they are excellent at putting cables (=fiber) in the ground.
As long as taxpayer funded entities put cables in the ground, it is necessary that these are available to all players to rent. This has the added value that the long write-off of these kind of cable systems, can be handled better by these than by smaller companies that have problems with cash-flow, who instead can concentrate on lighting up the fiber and offering services.
Ok, let's say we start giving people 10 or 100 meg internet access (this is already happening all across the world). When you start to aggregate thousands of these users, you really need 10Gig at least. How much do these people want to pay for their bandwidth? As little as possible.
Therefore the problem currently is not access speed or even intra-metro speed, it's the core long haul speed that is being crunched, or at least the cost of it. Another poster referred to $120k as being a "mere" cost for OC192 interfaces. This cost is WAY WAY too high (now, the list price of OC192 cards for the Cisco GSR is $225k per card, but that's another matter). Transporting bits long distances right now is cheaper than production cost because of the 2000 bubble buildout still hasn't run out of capacity, but we're getting there now. The cost of bw right now (can be as low as $15-25 per month per megabit of capacity) doesn't pay for the manpower and interfaces to move the bits long distances.
10Gig is not enough for the core, 40gig for SONET is getting close but won't be enough either, soon.
Already at 10gig you're running into all kinds of optical problems such as dispersion and so on, which needs to be handled. At 100Gig it's going to be quite a lot more problematic, if possible at all (you need to go 60-80km without repetition).
100meg internet access is enough for users, 10gig is enough for intra-metro, it's the inter-metro and international bw I'm worried about, and access speed isn't of as much use if you can only get good speeds within the same city as you're in.
I have been at RH7.3 since it came out and it works very well for me. I used to pay the $60 for redhat up2date support and thought that worked very well. I wish Redhat would have continued supporting it.
I was about to upgrade to Fedora Core 1 when I found out about the fedora legacy project which I think is a very good initiative.
The community driven initiative seems to be lacking support though, for instance the openssl updates have been in "testing" for 4-5 weeks now and still hasnt made it into the released-pool of updates. Being free I know I cannot demand anything, but I can observe that it doesnt seem to be working as well as I thought.
I'll probably go to Fedora Core 2 when it's released, it'd be nice to get the 2.6 kernel.
Slashdot Mirror
Another point to make is that at 8 megabit/s (not that uncommon speed for HSPA), spending those UKP240,000 takes ~11 hours. 40GB of data is approximately what you can fit on a standard Bluray disk.
That is what encryption is for. Get truecrypt or other similar application and then the data won't be extractable by anyone without the password.
The problem with ECN is that no major core router platforms support it, neither does the smaller routers. The only "core" networking device I know of that supports ECN is the CPU type routers (7200 et al) that Cisco makes. Their 12000 and CRS-1 do not.
There is little reason to have ECN turned on at end systems, if the intermediate devices that move/buffer packets doesn't set/use the ECN flag.
I agree. Flow routing with 4M flows didn't work in devices engineered in the late 1990ies, and they won't work now. Any script kiddie can create new "flows" by sending random port/dst IP packets thru the device, and it'll fall over and die just like the devices 10 years back did.
We stopped doing flow routing for a reason, it didn't work. Routers need to care about IP addresses and perhaps take into account port numbers to do load sharing between equal cost links, but nothing else. Looking into flows does NOT scale.
In Sweden, typical monthly peak average is between 20 and 400 kilobit/s depending on type of access and type of users.
This is equivalent to around a few GB per month (remember, it was peak bw usage during the day in 5 minute interval) to 40-60 GB per month.
This means that in most markets, 250GB is hit by a few percent of the users, but on the other hand 250GB per month is 0.7 megabit/s average usage, and with 10/10 megabit connections, a user can theoretically hit 10 times that cap, thus I understand why Comcast wants to do this.
Well, older cars are not protected of course (so there's your 50% right there) plus that an immobilizer doesn't protect against someone breaking into your car to steal stuff. It only means it's much harder to steal the actual car.
It became a requirement in all of EU in 1998 that all new cars must have immobilizer fitted from the factory to be allowed to be sold. Just weird that it's not a requirement in north america.
If you're running four full-bgp VPN-customers in your core routers that can handle 2M routes, you've done a major design mistake and that's not the fault of the hardware manufacturer.
Regarding routing table growth, hopefully IPv6 might stifle that a bit as we're going to be running out of IPv4 space in the next 3-5 years and IPv6 space is allocated in much larger blocks requiring fewer routes.
The CRS-1 is tested with at least 2M IPv4 routes. It should be enough to 2015 or more.
What some might have missed would be the opportunity to sell gfx cards due to all of a sudden being able to do H.264/VC-1 offload/acceleration in linux for HTPC usage.
The first manufacturer to offer this would get a lot of sales due to people using their cards for HTPC. I know I would, because currently I can't play 1080p VC-1 and barely h.264 properly, even on a Core2Duo.
Net Neutrality is not the answer to the problems seen in the US. The correct answer is to make the largest players rent out their infrastructure with bitstream access and LLUB (Local Loop UnBundled).
As soon as other companies can buy access to the customer and sell them services, then the largest players can't offer degraded or bad service, because the customer can go elsewhere. The problem that Net Neutrality tries to solve is a problem because the customers in a lot of areas don't have many companies to choose from. Solve that problem instead of trying to enforce Net Neutrality and the US will be much better off.
Sweden has had a system like this since 1960 or so. Whenever I do anything I give them my social security number (equivalent) and that's all I need. You cannot do anything in Sweden without one, unless you're all cash.
If I pay with a credit card I have to show ID. Identity theft is extremly rare here, I never even consider it. Very little personal business involving identity is done without showing for instance a drivers license (which also contains the Personal Identity Number).
The UK system of showing a gas bill or alike is just weird for someone like me.
Ok, one doesn't build residential broadband like ones builds T1 circuits to business. Everything you know about that you can just scrap when it comes to residential capacity.
Residential broadband over fiber is built using (ten)gigabit ethernet over dark fiber or CWDM at the distribution, and using 10/100 ethernet over fiber for the access fiber. This is the only way to make it low-cost enough for households to be interested.
Fiber to the county is fairly straight-forward.
Sweden is a country the size of californa with 9 million people. It is estimated that the cost of supplying fiber to within 30 meters of each household in Sweden costs approx $10bn, and if you skip the last 10% of households, you halve that cost to $5bn.
This is just dark fiber, not anyone actually lighting it up. But then again, comparing it to what roads cost per year, it's not very much money.
Here in Sweden we have a lot of municipality owned power companies that have entered the broadband business, some supplying fiber to the curb, some going all the way and connecting private houses and offering services.
Generally, these entities have little clue on what is important when offering services, whereas they are excellent at putting cables (=fiber) in the ground.
As long as taxpayer funded entities put cables in the ground, it is necessary that these are available to all players to rent. This has the added value that the long write-off of these kind of cable systems, can be handled better by these than by smaller companies that have problems with cash-flow, who instead can concentrate on lighting up the fiber and offering services.
It'll grow by the modular 3U unit.
The single 3U unit won't hold 67.2Tb, that's a bunch of them linked together.
Ok, let's say we start giving people 10 or 100 meg internet access (this is already happening all across the world). When you start to aggregate thousands of these users, you really need 10Gig at least. How much do these people want to pay for their bandwidth? As little as possible.
Therefore the problem currently is not access speed or even intra-metro speed, it's the core long haul speed that is being crunched, or at least the cost of it. Another poster referred to $120k as being a "mere" cost for OC192 interfaces. This cost is WAY WAY too high (now, the list price of OC192 cards for the Cisco GSR is $225k per card, but that's another matter). Transporting bits long distances right now is cheaper than production cost because of the 2000 bubble buildout still hasn't run out of capacity, but we're getting there now. The cost of bw right now (can be as low as $15-25 per month per megabit of capacity) doesn't pay for the manpower and interfaces to move the bits long distances.
10Gig is not enough for the core, 40gig for SONET is getting close but won't be enough either, soon.
Already at 10gig you're running into all kinds of optical problems such as dispersion and so on, which needs to be handled. At 100Gig it's going to be quite a lot more problematic, if possible at all (you need to go 60-80km without repetition).
100meg internet access is enough for users, 10gig is enough for intra-metro, it's the inter-metro and international bw I'm worried about, and access speed isn't of as much use if you can only get good speeds within the same city as you're in.
I have been at RH7.3 since it came out and it works very well for me. I used to pay the $60 for redhat up2date support and thought that worked very well. I wish Redhat would have continued supporting it.
I was about to upgrade to Fedora Core 1 when I found out about the fedora legacy project which I think is a very good initiative.
The community driven initiative seems to be lacking support though, for instance the openssl updates have been in "testing" for 4-5 weeks now and still hasnt made it into the released-pool of updates. Being free I know I cannot demand anything, but I can observe that it doesnt seem to be working as well as I thought.
I'll probably go to Fedora Core 2 when it's released, it'd be nice to get the 2.6 kernel.