March 16th, 2008
P4P: faster, smarter P2P
The P4P working group demo’d higher P2P download speeds with 1/6th the inter-metro hops that soak up expensive, long-distance network bandwidth. P4P is designed to enable better ISP and P2P coexistence with a win/win solution: better performance for users and less network overhead for the ISP.
P4P speeds up P2P downloads by localizing network traffic. Standard P2P traffic gets spread all over the globe, so a single packet may go through a half-dozen costly high-end routers and thousands of miles of scarce ocean-floor fiber on its way to your PC. Metro-area routing is both cheaper for ISPs and faster for the users.
[graphic courtesy of the P4PWG]
Traditional P2P
Today’s P2P is network oblivious: peers are selected without regard to network topology. One stream may be coming from Flagstaff while the next is coming from Cape Town.
pTracker and iTracker
P4P is an open standard for delivering network awareness to P2P networks. One way it can work - and there is more than one, but the tech papers aren’t available yet on the web - is to add a peer-tracker (pTracker) and an Internet-tracker (iTracker).
The peer queries the pTracker for nearby peers. If the pTracker knows where the right bytes are in the local metro area, it returns that info to the peer. But if the inventory isn’t locally available, the pTracker could then go to the iTracker for peering suggestions that take into account network topology and costs.
The pTracker then selects a set of active peers and returns that list to the peer that initiated the request.
The pTracker is run by the P2P system - say BitTorrent or Pando - while the iTracker can be run by trusted 3rd parties, P2P networks or ISPs.
The Storage Bits take
As the volume of video content rises, the demand for P2P will only grow. With P4P it appears a substantial portion of the ISP community will make its peace with P2P content distribution. This is a Very Good Thing.
Which leaves Comcast - who isn’t a P4PWG member - facing their original problem: they under-provision their network in order to maximize profit. As long as customers are passive receptacles that works, but once they start distributing content through P2P Comcast has a problem.
Comcast wants to make their problem your problem. Rather than saying they can’t compete with DSL or fixing the problem through protocol or equipment upgrades, they’ve been fighting the common-carrier law.
That’s just wrong. Common carrier status for telecom is over 160 years old. It has stood the test of time for very good reasons. Comcast needs to get with the program: either get competitive with the telcos or get out.
Comments welcome, of course.
Robin Harris has been messing with computers for over 30 years and selling and marketing data storage for over 20 in companies large and small. See his full profile and disclosure of his industry affiliations.
Subscribe to Storage Bits via Email alerts or RSS.
| 03/17/08
