August 12th, 2012
As more and more organizations have started to use the cloud for their computing needs (I have been using Amazon Web Services, commonly known as the EC2 cloud for more than two years now), a relatively new set of challenges has arisen. The cloud providers provide computing resources, while the organizations care about their analytical and computing jobs being completed, irrespective of the computing resources they require. The organizations would like to place a value on the job, instead of the resource, and there is currently a missing link between the two.
As a regular reviewer for Computing Reviews, I have just finished a formal review for Near-optimal Scheduling Mechanisms for Deadline-Sensitive jobs in Large Computing Clusters, and that work specifically tries to address these new kinds of challenges that we are all observing as the movement to cloud computing gathers further steam.