Service Reliability Engineering - Failures are Always an Option


Sep 16, 2018

Modern software systems are usually built using smaller, interconnected components. The rise of the microservices architecture popularity tends to increase the overall number of components compared to the traditional monolith approach. Both components and interconnections are subject of failure. The risk of failure also increases with the number of components, decreasing the overall system reliability.

The talk aims to outline the most common sources of failure and to provide engineer’s consciousness about making decisions on how to design, deploy and monitor such systems and ultimately increasing the reliability of the service. The talk balances a mix of distilled industry giants’ experience as well as hands-on observations.