Closing keynote: Above the clouds - A view from academia
Since becoming a commercial reality in late 2007, cloud computing has shaken up every area of information technology, and appears to be anything but a passing fad. As early academic adopters of cloud computing, we quickly encountered important questions about its uses in academia.
What research and teaching should be done in the public cloud vs. our own systems?
Should UC Berkeley consolidate and "cloudify" its existing computing facilities?
Should the University of California system - nine major research universities - build and operate a UC-wide "academic cloud"?
What about non-commercial cloud efforts in the US, such as the OpenCirrus platform, the CluE cluster operated by IBM and Google and supported in part by the National Science Foundation, and Yahoo's M45 cluster for running Hadoop jobs - will there be a separate "academic cloud" market or will it be co- opted by existing public cloud providers?
I'd be crazy to try to prognosticate the answers, but I hope to stimulate discussion based on our experience at Berkeley.