In the world of Kubernetes, running stateful workloads like PostgreSQL databases often raises concerns about stability and resilience. This session aims to address these concerns head-on through resilience and chaos testing. I'll delve into the robustness of PostgreSQL operators by subjecting them to extreme conditions: killing Pods, losing Persistent Volume Claims (PVCs), and more. Our goal is to observe how well these operators recover and the impact on downtime for client applications. This talk is about facts, evidence and 2 operators:
- CloudNative PG
- Patroni-based operator
Join me as I push PostgreSQL operators to their limits and uncover the truth about their reliability under chaos.