Overview

As in so many other domains, data-driven research is becoming an indispensable way of developing solutions for distributed computer systems. This way of working is shaping the next generation of systems researchers. A typical example is using machine-learning techniques for analysis of network traffic for reasons of security: a researcher will have to understand the tools, while at the same time being an expert on security. Likewise, closed-loop architectures are being developed for monitoring cloud computers and predicting their performance, where predictions are based on, again, machine-learning techniques . Finally, to understand massively decentralized communication structures, one often uses techniques from complex-network analyses. These examples illustrate that the field of distributed systems touches upon much more than what we tend to think. In this brief talk, I will focus on some of these examples with the intent to inspire systems researchers to explore tools and techniques they would normally not easily adopt.