Keynote: Distributed Systems Simplified Using Kubernetes

on YouTube ↗06m

" summary

Explains an etcd operator that encodes human ops logic to manage distributed etcd clusters on Kubernetes, using an observe–analyze–act loop and calling the Kubernetes API. Demo runs a three-node cluster, destroys one node’s disk state, and the operator recreates a replacement pod and restores cluster membership so the cluster remains available. Covers rolling upgrades, backups, integration with the Prometheus operator via ServiceMonitor, and points to the open‑source etcd-operator on GitHub.

AI-written summary · about

" chapters (10)

1. 0:00Start
2. 0:49ReplicaSet
3. 1:04Creating a Database is Easy on kubernetes
4. 1:24Managing a Distributed Database is Harder
5. 2:02The Dream
6. 2:39etcd Overview
7. 2:56etcd Operator Resource
8. 4:29etcd Operator - Future Work
9. 4:39Prometheus Operator
10. 5:24Next Steps

" related talks

• 24mAutomating Stateful Apps with Kubernetes OperatorsJan Kleinert · DevOpsDays Boston 2019
• 36mKubernetes Day 2: Cluster Operations [I]Brandon Philips · CloudNativeCon + KubeCon Europe 2017
• 05mLet’s Deploy etcd OperatorArka Saha · KubeCon + CloudNativeCon Europe 2026
• 53mAutomating Stateful Applications with Kubernetes OperatorsGraham Dumpleton, Jorge Morales · FOSSASIA Summit 2019
• 31mDemystifying Kubernetes Failure Scenarios Caused by etcdWilliam Caban · KCD Guatemala 2023
• 37mDeep Dive: kops (Cluster Lifecycle) SIGJustin Santa Barbara, Mike Splain · KubeCon + CloudNativeCon Europe 2019
• 39mBuilding Distributed Systems with KubernetesErik St. Martin · GOTO Chicago 2018
• 18mKeynote: What is Kubernetes?Brian Grant · KubeCon + CloudNativeCon North America 2017