Lessons from a Kubernetes Cluster Upgrade
Upgraded our staging clusters to v1.19 this quarter.
Lessons learned:
- Always check deprecations beforehand (
kubectl deprecations) - Test ingress controller compatibility early
- Pod disruption budgets saved us from some flakiness
Tooling we used:
kustomizefor manifestsvelerofor backups
Planning is everything. Production upgrade next quarter!
This post is licensed under CC BY 4.0 by the author.