Page One

Campus Computing News

Designing UNT Websites for Accessibility

Lab-of-the-Month: Willis

Today's Cartoon

RSS Matters

SAS Corner

The Network Connection

Link of the Month

WWW@UNT.EDU

Short Courses

IRC News

Staff Activities

Subscribe to Benchmarks Online
    

Network Connection

By Dr. Philip Baczewski, Associate Director of Academic Computing

Peer to your Peer

The next "big thing" on the Internet may be already in the works. Peer to peer network applications have already had an impact in some circles, but that may only have been the tip of the iceberg. Peer to peer is written in Internet shorthand as p2p. One p2p application you may have heard of is Napster. Others exist or are being developed as you read this column. P2p applications potentially can tap into the enormous underused pools of storage, processing, and bandwidth resources that exist in those millions of personal computers around the world.

Nap it in the bud?

Napster made a big splash by allowing people on the Internet to trade music files. The way it worked was that the files were stored on the computers of the people that used Napster. When you used Napster to download files, you were actually retrieving them from other Napster users computers. Any files you retrieved were made part of the Napster "network" and would be made available to other Napster users who wanted them and happened to be close, network wise, to you. In other words, your computer formed a network with its "peers" to store and provide music files via the Napster application. The reason that the RIAA could shut Napster down via copyright lawsuits was that the company kept a central database of the Napster "peers" and the music files they stored. This was deemed to be facilitation of violation of the copyright law and Napster inc. has paid the price.

Drawing the ire of the RIAA was not the only impact Napster had. It turned out that the p2p design was quite an effective way to provide people access to data on the Internet -- so effective, in fact, that it caused enormous amounts of bandwidth to be absorbed just by that application alone. This was particularly true on college campuses, where there exists a large population of people interested in popular music with access to high-bandwidth computer networks and a proclivity for using new technology. Most college campuses (including ours) responded by simply throttling back the bandwidth that Napster traffic could use. What stopped the traffic, however, was the RIAA's success in pounding the company to a pulp with its legal sledgehammer.

Its not Just a Napster Thing...

Napster was not the only p2p application to gain attention in the last year or more. Gnutella is a program similar to Napster which allows sharing of files and has been used extensively to trade music. Last fall, at times over 25 percent of Internet2 traffic could be attributed to a program named KaAaA, which not only supports the trading of music files, but also of video files. Obviously, there's something about the p2p concept which has made it a popular venue for exchanging information, and in particular media files over the Internet.

P2p applications can take advantage of a few concepts which have made them amenable to the types of activity for which they have been used so far:

  • they don't rely on one central server and remove any bottleneck that a single source can create

  • they can tap multitude of sources and use the one which is closest, network wise, to the download point

  • they can grow on their own without the need of a central organizing authority

  • they can do their work anonymously if designed to do so

P2p applications don't necessarily use network bandwidth efficiently and thus can be the bane of network administrators everywhere.

What's to Come?

P2p applications by themselves are not useful until they support an activity that people really want to do. A lot of people, whatever their motivation, wanted to share music files or at least be able to download and listen to a lot of music without having to spend their fortune in records stores. As more types of activities are discovered that can benefit from p2p, we can expect to see more applications on the horizon. If you are more interested in reading more about it, I suggest the following: http://www.oreilly.com/www/oreilly/press/p2presearch.html. P2p: it's out there and more is coming.