Abstract:

With the arrival of cloud computing and the Internet-of-Things, programming is growing increasingly distributed. And yet, anyone who attempted to build a complex distributed system is well aware that distributed programming is exceptionally hard.

The goal of the reliability-oriented computing approach is to significantly simplify the construction of distributed systems by treating reliability as a first-class citizen.

Distributed programming is simplified from several angles, from encapsulating the essential distributed algorithms into reusable building-blocks, exposing type systems that check the correctness of distributed programs, to providing a safe, transparent execution runtime for the IoT network.

In this talk, I will start by explaining some basic principles of concurrent programming. I will then draw some parallels between concurrent and distributed computing. Finally, I will explain the three pillars of reliability-oriented computing -- the programming model, the type systems and the execution runtime, and show how they enable creating more robust distributed systems.