Harnessing the Power of Chaos Engineering

[sh-aps]

2020 was a bit chaotic, as we all know. For some retailers, it meant speeding up digital transformation. We all knew it was going to be a bumpy road but were we ready? Big spoiler: we were not ready for a lockdown scenario. We knew we were going to fail but we needed to measure how fast we could recover. How did we do? Monitoring system to detect anomalies Mix of web apps and Kubernetes with refactors ongoingContinuous performance testing in a pre-production environment Chaos Testing in productionWhat did we want to test? Scalability and resilience, load spikes like Black Friday, Christmas Campaigns, Flash Offers, and more as well as system failures such as pod failure and network problems. In this talk, we will explain how we implemented Chaos Testing in our production environment in Azure with Chaos Mesh, k6 and Litmus. We will be showing different examples like availability in different regions, database access problems, or Kubernetes different chaos scenarios.