Table of Contents | ||
---|---|---|
|
This is adapted from: https://cluster-api.sigs.k8s.io/tasks/upgrading-clusters.html
...
It’s recommended to regularly upgrade your clusters. This avoids trying to maintain tooling version compatibility against major Kubernetes versions.
OS Patches
The operating system and associated packages can be updated independently, e.g. to apply security patches to the host OS.
The Ubuntu image is stripped so the packages (and number of vulnerabilities) is significantly lower. The Cloud team will make it clear when a CVE applies to the CAPI Ubuntu images.
Note |
---|
Containers are controlled by Kubernetes. If a container (e.g. |
Multiple Version Upgrades
...
To upgrade major versions you will need to follow the https://stfc.atlassian.net/wiki/spaces/CLOUDKB/pages/285704256/Cluster+API+Upgrade#Upgrade-Clusterctl-and-CAPI-components section first then https://stfc.atlassian.net/wiki/spaces/CLOUDKB/pages/285704256/Cluster+API+Upgrade#UpgradingUpgrade#Kubernetes-KubernetesImage-Majorand-Version-Upgrades for each hop.
Overview
This process assumes the administrator is doing a full upgrade of all components. These can be upgraded independently with the caveat that the Infrastructure layer supports the version of Kubernetes planned: https://cluster-api.sigs.k8s.io/reference/versions
...
Update the helm Cluster API charts:
Info |
---|
The upstream repo URLs changed during Q3 2024. You need to update the URLs if your |
Code Block |
---|
# Update URLs to repos if required: helm repo add capi https://azimuth-cloud.github.io/capi-helm-charts --force-update helm repo add capi-addons https://azimuth-cloud.github.io/cluster-api-addon-provider --force-update # Avoid updating openstack-cluster chart until later helm repo update capi-addons helm upgrade cluster-api-addon-provider capi-addons/cluster-api-addon-provider -n clusters --wait cd <folder_with_values> |
...
Validate that the upgrade is valid and apply the command provided by clusterctl
Kubernetes Image and Version Upgrades
This section assumes production clusters and upgrades components individually.
For development / low risk clusters both steps can be combined into a single roll-out.
Without minor version upgrade
...
You’ll need to upgrade VM images and Kubernetes version to the latest patch version available
...
before doing any major upgrades:
I.e. if you’re on 1.100.12 upgrade to the latest 1.100.x, this ensures any bug-fixes are applied which could prevent later upgrades.
For major and minor
...
upgrades
...
:
Lookup the latest image build for Kubernetes, this can be found in images section of the web interface
Edit the
kubernetesVersion
inuser-values.yaml
...
to
...
match the image name
Edit the
machineImage
inuser-values.yaml
...
to
...
use the latest patch release
Code Block |
---|
helm repo update capi
helm upgrade <cluster_name> capi/openstack-cluster -f values.yaml -f clouds.yaml -f user-values.yaml -f flavors.yaml -n clusters |
Wait for the rollout of new infra to complete
The rollout can be monitored with
kubectl get kcp -A
andkubectl get md -A
Machine details can be found in
kubectl get machines -A
andkubectl get openstackmachines -A
With minor version upgrade
- Upgrade to the
Repeat for each major upgrade step
You can only do a single major upgrade at a time, e.g.
1.100.12
to1.101.4
, then1.102.6
Supported Kubernetes Versions
We aim to support Kubernetes versions and keep CAPI images for N -1 → all supported End of Life (EOL) versions from: https://kubernetes.io/releases/ OR the oldest CAPI image we have that gives an upgrade path - whichever is the older; where N is the latest Kubernetes version.
For example, as of the 29/01/2025, the latest Kubernetes version is v1.32 and therefore, we will provide CAPI images for Kubernetes versions v1.31 → v1.26 (oldest image we hold).
Once we provide the latest N-1 image, we will keep the oldest image (v1.26) for a minimum of 4 weeks to let you upgrade older clusters and notify any users accordingly.
Please refer to the Kubernetes Release page for the latest version history and patch notes.
Troubleshooting
On the management cluster
...