In the just concluded INFORMS Annual Meeting 2011 at Charlotte, NC, I attended an interesting session on a survey of optimization problems in cloud computing. There were actually a bunch of talks on cloud computing, definitely more than what I noticed in the last annual meeting in 2010. That, of course, closely correlates with the importance and popularity of cloud computing, and more vendors entering this space.
The survey was very well presented. One aspect I pointed out to the speaker was the somewhat absence of stochastic optimization problems in this domain. I think quite a few of the problems mentioned in the talk are actually stochastic in nature, but were presented as deterministic mostly for simplicity.
I encourage anyone interested in this area of research to get in touch with the presenter for the slides.