Abstractions, Models and Tools for Scalable Systems: Progress and Open Problems

The growth of mobile and cloud computing, cyberphysical systems and the internet of things has arguably made scalable ("Big Data") concurrency the dominant form of computing. I will describe the state of the art in models of concurrency, programming abstractions and analysis tools. I will then summarize how actor languages and frameworks have been widely adopted to address scalability, and how new analysis tools that are facilitating software safety. However, as we scale up, a key limitation of current languages and tools becomes apparent: the difficulty of representing quantitative and probabilistic properties and reasoning about them. I will conclude by discussing problems that need to be addressed to simplify building robust scalable concurrent applications.