Day 2 Operations of Helm Charts
summary
Argues Kubernetes YAML mixes app and runtime concerns so templating fails for third‑party charts and proposes Helm + customize as a patch‑not‑fork workflow; demonstrates a tool called KOTS that templates upstream charts, stores small strategic patches per environment, and opens Git pull requests so chart updates flow into GitOps without manual merges. Claims merge conflicts vanish in 90% of cases and shows a working demo creating staged and prod PRs and applying updates.
AI-written summary · about
chapters (9)
- 0:00Start
- 3:08limited deployment environments
- 7:18fork the chart
- 19:17install a new application
- 19:48change the replica count to two
- 20:04add additional customizations
- 20:19add in a nic container
- 21:17add deployment to a cluster
- 23:02adding a parameter for the log level
related talks
43mAt The Helm of Kubernetes: Repeatable Infrastructure Creation for Mere Mortals
34mKustomize vs Helm - Civil War
41mHelm - the Better Way to Deploy on Kubernetes
10mUsing Helm to the Fullest
29mHelm Chart Patterns
27mIntro: Helm
32mThe Automation Challenge: Kubernetes Operators vs Helm Charts
26mHelm your way with Kubernetes