DevOps for Mere Mortals
summary
Large org runs custom CI with Jenkins and internal tools that handle 100k–200k builds per day; small startup of ~5 engineers builds a lightweight pipeline using Travis + GitHub + Artifactory, Docker images from service repos and a separate service and deployment repos pattern. CI jobs push images, auto-bump versions, commit updated Kubernetes YAML into a deployment repo, then trigger rolling updates to GKE. Pain points include shared-infra outages, leaked secrets, and Java-in-container OOMs from missing cgroup flags; the demo robot commits and deploys successfully.
AI-written summary · about
chapters (5)
- 0:00Start
- 0:27Introduction
- 2:43"DevOps in the Small": SADA Systems
- 6:26From Commit to Deployment
- 26:40Why change what you are doing?
