Kubernetes frequently gets patches and upgrades, addressing security vulnerabilities and adding new functionality. Additionally, AKS releases patches for the underlying node operating system on a regular basis (aka Node Image Updates). You need to implement a process to upgrade the cluster.
Make sure the following are complete before setting up ingress.
- Cluster is provisioned and accessible via 'kubectl'
- App Deployment is complete
- Initial testing should use a manual upgrade process
- The UI running pod count should be increased to 2
- The UI pod count should never go below 1 during an upgrade
- Day 1 (simulated): Due to a critical OS level CVE you've been asked to upgrade the system pool NODE IMAGE ONLY
- Day 2 (simulated): Due to a critical Kubernetes level CVI you've been asked to upgrade the control plane and the system pool Kubernetes version to the next incremental version (major or minor)
- Day 3 (simulated): To take advantage of some new Kubernetes features you've been asked to upgrade the user pool Kubernetes version to the next incremental version (major or minor)
-
Increase the Red Dog UI deployment replica count to 2
-
Deploy the necessary config to ensure the UI pod count never dips below 1 pod
-
Check the available upgrade versions for Kubernetes and Node Image
-
Upgrade the system pool node image
-
Upgrade the AKS control plane and system pool Kubernetes version
-
Upgrade the user pool Kubernetes version
Bonus Tasks:
-
Enable Automatic Upgrades to the 'patch' channel and set a Planned Maintenance Window (preview) for Saturdays at 1am
Useful links: