Epidemic Computing at Cornell

http://weblogs.cs.cornell.edu/AllThingsDistributed/archives/000456.html

In a posting last week I gave a bit of history of the use of Epidemic techniques in computing and why we are so interested in them when it comes to distributed computing. It is remarkable that many of the techniques that make systems scalable in general, are at the a part of epidemics. Of course biological epidemics scale really well and are very robust: once a few subjects are infected it is almost impossible to stop the spread, even if you isolate the original source.

One of the other cool properties of epidemics that I didn't mention in the original posting is they work better and better at large scale. This means that having to scale up your system is no longer a burden, but it becomes an advantage and you can deliver on the true promise of scale-out: more nodes means a more robust system.

Various groups at Cornell have been looking at problems using similar techniques. For example Jon Kleinberg and David Kempe's work on gossip protocols & small world phenomena, spread of influence in social networks, or David's work with Alin Dobra and Johannes Gehrke on computing aggregate information using gossip. And of course since the return of Alan Demers we have on of the founding father of epidemics techniques in house