3 ms·

I still don't understand why this is a terrible idea. Is there absolutely no way for ISPs to react to this in such a way that they save money rather than lose i

I still don't understand why this is a terrible idea. Is there absolutely no way for ISPs to react to this in such a way that they save money rather than lose it? If decreasing traffic between providers doesn't save money, then wouldn't there be some savings at least from decreasing traffic between subnets?

brazzy 17y ago
CDNs don't scale down that well. The likelihood that there is demand for one particular piece of data within the same subnet (and thus the value of having it there) decreases dramatically with the size of the subnet. As the article itself points out, some CDNs already do this with large ISPs, directly. This kind of thing may increase and get more sophisticated, but it will never include the users themselves as middle men.
- milkshakes 17y ago
  CDNs don't scale down that well. The likelihood that there is demand for one particular piece of data within the same subnet (and thus the value of having it there) decreases dramatically with the size of the subnet. Is there no inflection point in between, where this makes sense? I guess what I'm trying to say is that CDNs cant be everywhere, but users can
- marcamillion 17y ago
  I agree with milkshakes on this one. Brazzy...you are right. CDNs (in their current form) don't scale well. A properly executed p2p CDN can scale much better, IMHO. As for not including users themselves as middle men, I must disagree with that. It makes economic sense for the ISPs to want to convert bad p2p traffic to good p2p traffic. That's their goal...all the while monetizing that good p2p traffic. Put it this way, the ISPs will do nothing, except sign a revenue share agreement and possibly change their TOS to allow customers to resell their bandwidth. For every GB resold, the ISP makes money. So in essence, they are converting bad p2p traffic to good p2p traffic and making a killing - the economics works in their favor.
  - brazzy 17y ago
    To the ISP there is no "bad" or "good" P2P traffic, just costs vs. income. And why would an ISP want to share revenue on bandwidth rather than get all the revenue themselves, since they are ultimately providing it?
jerf 17y ago
"Is there absolutely no way for ISPs to react to this in such a way that they save money rather than lose it?" Yes, the one I laid out. The ISPs ban their customers from participating for "bandwidth reasons", then directly provide the services to the distributed CDN, rather than going through the users. My argument is not that the ISPs will find this too expensive; it's the exact opposite! If there is value here, the ISPs will find it too profitable to resist, and they will not let their users in to the party. The problem here is that in the long run, the users aren't bringing anything to the party; they only bring bandwidth, which isn't actually theirs, it belongs to their ISP. (And some fairly minimal storage, but that is already a commodity and rapidly becoming even more so; CPUs have stalled some but all the other bits of the computer are still marching on Moore's style.) For contrast, consider a "legal botnet" that does computational work where the computation dominates the data transfer. We know that works; SETI@home has been running for over a decade. Users simply divvy up the ISP's bandwidth, but users own orders of magnitude more processing power than the ISP does, so there's actually a resource here that the ISP can't just tap itself. (Well.... legally anyhow.) I'd even say the fact that the ISPs have not already implemented this distributed CDN is a powerful argument against the idea that there's a lot of value here to be captured, though of course that's not proof.
- bdonlan 17y ago
  Actually, akamai hosts their CDN nodes right at ISPs already - though this is mostly to serve customers on that very ISP without paying for backbone bandwidth. So in a way we already have this distributed CDN...
  - jerf 17y ago
    Yeah, I thought that was the case but wasn't 100% sure, so I didn't mention it. This further dilutes the value of a botnet CDN, probably straight into negative territory.

brazzy 17y ago

CDNs don't scale down that well. The likelihood that there is demand for one particular piece of data within the same subnet (and thus the value of having it there) decreases dramatically with the size of the subnet. As the article itself points out, some CDNs already do this with large ISPs, directly. This kind of thing may increase and get more sophisticated, but it will never include the users themselves as middle men.

milkshakes 17y ago
CDNs don't scale down that well. The likelihood that there is demand for one particular piece of data within the same subnet (and thus the value of having it there) decreases dramatically with the size of the subnet. Is there no inflection point in between, where this makes sense? I guess what I'm trying to say is that CDNs cant be everywhere, but users can
marcamillion 17y ago
I agree with milkshakes on this one. Brazzy...you are right. CDNs (in their current form) don't scale well. A properly executed p2p CDN can scale much better, IMHO. As for not including users themselves as middle men, I must disagree with that. It makes economic sense for the ISPs to want to convert bad p2p traffic to good p2p traffic. That's their goal...all the while monetizing that good p2p traffic. Put it this way, the ISPs will do nothing, except sign a revenue share agreement and possibly change their TOS to allow customers to resell their bandwidth. For every GB resold, the ISP makes money. So in essence, they are converting bad p2p traffic to good p2p traffic and making a killing - the economics works in their favor.
- brazzy 17y ago
  To the ISP there is no "bad" or "good" P2P traffic, just costs vs. income. And why would an ISP want to share revenue on bandwidth rather than get all the revenue themselves, since they are ultimately providing it?

jerf 17y ago

"Is there absolutely no way for ISPs to react to this in such a way that they save money rather than lose it?" Yes, the one I laid out. The ISPs ban their customers from participating for "bandwidth reasons", then directly provide the services to the distributed CDN, rather than going through the users. My argument is not that the ISPs will find this too expensive; it's the exact opposite! If there is value here, the ISPs will find it too profitable to resist, and they will not let their users in to the party. The problem here is that in the long run, the users aren't bringing anything to the party; they only bring bandwidth, which isn't actually theirs, it belongs to their ISP. (And some fairly minimal storage, but that is already a commodity and rapidly becoming even more so; CPUs have stalled some but all the other bits of the computer are still marching on Moore's style.) For contrast, consider a "legal botnet" that does computational work where the computation dominates the data transfer. We know that works; SETI@home has been running for over a decade. Users simply divvy up the ISP's bandwidth, but users own orders of magnitude more processing power than the ISP does, so there's actually a resource here that the ISP can't just tap itself. (Well.... legally anyhow.) I'd even say the fact that the ISPs have not already implemented this distributed CDN is a powerful argument against the idea that there's a lot of value here to be captured, though of course that's not proof.

bdonlan 17y ago
Actually, akamai hosts their CDN nodes right at ISPs already - though this is mostly to serve customers on that very ISP without paying for backbone bandwidth. So in a way we already have this distributed CDN...
- jerf 17y ago
  Yeah, I thought that was the case but wasn't 100% sure, so I didn't mention it. This further dilutes the value of a botnet CDN, probably straight into negative territory.