The Container Operator’s Manual
summary
Says containers work well for ephemeral stateless services and testing but fight with stateful databases — advice: prefer cloud DBs or keep containerized databases tiny, because network‑bound setups and per‑host volumes add cost and complexity. Cites running Docker 0.6 in Jan 2015, operating Dockerized Cassandra clusters, and running the Kubernetes platform behind github.com. Estimates platform rollout as ~one year and staffing of roughly team of 6–8 (minimum four), plus choices for orchestration, networking and monitoring.
AI-written summary · about
chapters (23)
- 0:00Intro
- 4:52What is a container
- 7:00Native features
- 14:52Are you Google
- 15:52Benefits
- 16:45Cost
- 17:25Network bound
- 18:12Containerized databases
- 19:29Keep it small
- 19:57Automatic Database Failover
- 20:32Containers Need Friends
- 21:05Container Orchestration
- 21:58Networking
- 22:26Networking Software
- 23:07Container Platform
- 25:56Hybrid Solution
- 26:44Containers need headcount
- 27:10Containers are like Jenkins
- 27:53Build a new team
- 32:42Team size
- 33:04Authority
- 35:00Should we use containers
- 36:09Do you want containers
related talks
38mThe Container Operator's Manual
25mAre Containers Ready for Production Databases?
47mFrom a Database in Container to DBaaS on Kubernetes
43mContainers Change Everything
01h 06mPanel: Running Databases in Containers
01h 02mContainers in Production: It’s Like Orchestrating Cats
36mDocker Inside/Out
36mFrom a Database in Container to DBaaS on Kubernetes