“Peer-to-peer has been around for a long time, but it is now being recognized as the computing paradigm of the future,” said Albert Yu, senior vice president of the Intel Architecture Group.
Not so fast, Mr. Yu.
While a number of P2P applications have recently captured the imagination of venture capitalists, technology barons and, of course, Wall Street, there are some serious challenges – and serious risks – to be found down that particular path. True, Napster, Gnuttela, Groove and other distributed applications offer nifty functionality. But let’s not forget some of the baggage that comes along with a P2P implementation:
Service broadcasts – any peer that offers services needs to make those services known to peers that would be clients. Those broadcasts chew up bandwidth. Dozens of peers make for a noisy LAN. Millions of peers could make an unusable Internet.
Directory services – Want to eliminate peer service broadcasts? Implement a directory service that acts as a sounding board for peer services. Of course, then the directory service itself becomes a significant point of failure and a potential bottleneck.
Today’s P2P applications do address the greatest traditional P2P obstacle… that is, finding what you’re looking for among a zillion distributed peers. At the same time they raise new issues: how does a user know if the content in question is authentic? Complete and unaltered? Virus-free? Legal to download? From a trustworthy source? From a source with usable download performance?
Let’s face it… server-based Internet computing has a lot going for it. For peer to peer services to take over, they’ve got lots of questions to answer… and a lot of growing up to do.