These are the old pages from the weblog as they were published at Cornell. Visit for up-to-date entries.

December 10, 2004

Worlds'04 & Real & Large

In yesterday's post about Worlds'04 I didn't intend to criticize the Worlds organization or research in general for having a different scope than my problem set of It is just that given the issues that are important for Amazon, there were only a few relevant presentations. It is not that I think the research performed is not relevant, as Thorsten von Eicken mentioned in his Keynote at Middleware, there are still many, many problems to be solved for systems with a hundred or more nodes.

As David remarks in the comments yesterday, general research does not have access to the problems of the Amazon's and EBay's of this world. That is true, but we are starting to change that. We are working on getting sabbaticals and graduate level internships set up, with the first people getting starting in these programs in January 2005. We are working hard on getting other exchanges going to facilitate sharing of ideas.

Posted by Werner Vogels at December 10, 2004 02:41 PM


That sounds great!

It's interesting, grad students traditionally do internships at the research divisions of "platform" companies like Microsoft, IBM, and HP, but there has been very little available (to my knowledge) at Internet services. (I vaguely recall Google having some grad student internships, but I think that started only very recently.)

I definitely agree with you that Amazon/eBay/Google/Akamai/... scale is a totally different environment than PlanetLab or the Grid, given that the scale is almost 1000x larger. Hopefully the sabbaticals, internships, and other collaborations you are working on will help start to bridge the gap between the relatively toy systems that academics have no choice but to work with, and the systems that are used in the "real world" !

Posted by: David Oppenheimer on December 10, 2004 05:19 PM