Kubernetes Operators: What are They and Why would You Need Them?
summary
Operators are Kubernetes controllers that run outside the control plane and encode domain knowledge to automate day‑two ops — observe the CRD/CR, analyze differences, act to reconcile (the control loop runs every ~30s or on events). Examples include the prometheus operator, database operators and Cluster API; tooling discussed includes kubectl, CRDs, Helm, Argo CD/Flux and KEDA. Operators behave like microservices: they create real resources, usually work, and require testing and ongoing maintenance.
AI-written summary · about
chapters (23)
- 0:00Intro
- 1:17Kubernetes Objects
- 4:06Custom Resource Definition
- 7:12Custom Resources
- 7:33How do operators work
- 7:49Controller pattern
- 9:33Operator definition
- 11:15What doesnt an operator do
- 14:28Prometheus Operator
- 16:16User Interaction
- 18:09Kubernetes Lifecycle
- 19:51Day 2 Operations
- 22:34Examples
- 23:20Database Operators
- 23:42Application Operators
- 24:57Should You Use Operators
- 26:40Should You Build an Operator
- 27:02Do You Really Need an Operator
- 30:31How to Build an Operator
- 31:03Debugging an Operator
- 31:45Debugging Checklist
- 39:24Conclusions
- 41:07Questions
related talks
45mKubernetes Operators: Operating Cloud Native services at scale
07mWhy You Care About Kubernetes Operators
21mOperators for Beginners
01h 58mKubernetes Operators Within Reach of Any Developer
31mKubernetes Operators within Reach of Any Developer
28mSmooth Operator - A Rough Guide to Kubernetes Operators - Olive Power, VMware
14mDemystifying k8s operators
20mSmooth Operator - A Rough Guide to Kubernetes Operators