Monitoring, alerting and log aggregation are essential for the smooth functioning of a production grade Kubernetes cluster. This article explains why they are necessary, the requirements for a production-grade logging system and how we have implemented the same at Halodoc.