Schedule this migration during a maintenance window or low-traffic period to minimize service disruption.
Migrating End-of-Life instance pools in your Kubernetes Kapsule cluster
Scaleway is deprecating support for certain Instance types that have reached their End of Life (EOL). This guide outlines the recommended steps to migrate your Kubernetes Kapsule cluster node pools from deprecated Instance types to currently supported ones.
Before you startLink to this anchor
To complete the actions presented below, you must have:
- A Scaleway account logged into the Scaleway console
- Owner status or IAM permissions allowing actions in the intended Organization
- Created a Kubernetes Kapsule or Kosmos cluster
Identifying deprecated Instance poolsLink to this anchor
- Log in to the Scaleway Console.
- Navigate to Kubernetes under the Containers section in the side menu of the console.
- From the drop-down menu, select the geographical region you want to manage.
- Select the cluster containing the node pools using deprecated Instances.
- In the Pools tab, check the Instance type column for any pools using deprecated or soon-to-be-removed types.
Creating replacement pools with supported Instance typesLink to this anchor
-
For each ENT1 pool identified:
- Click + Create pool (or Add pool).
- Choose a supported Instance type from the Node Type dropdown menu.
- Configure the pool settings (e.g., Availability Zone, size, autoscaling, autoheal) to mirror the existing pool configuration as closely as possible.
- Click Create (or Add pool) to initiate the new pool.
-
Monitor the status of the new nodes until they reach Ready state:
- In the Pools tab of the console.
- Alternatively, use
kubectl
with the command:kubectl get nodes
Migrating workloads to the new poolLink to this anchor
-
Cordon the deprecated nodes to prevent them from receiving new pods:
kubectl cordon <deprecated-node-name> -
Drain the deprecated nodes to reschedule workloads onto the new nodes:
kubectl drain <deprecated-node-name> --ignore-daemonsets --delete-emptydir-dataNoteThe flags
--ignore-daemonsets
and--delete-emptydir-data
may be necessary depending on your environment. Refer to the official Kubernetes documentation for detailed information on these options.
These commands ensure that your workloads are running on the new nodes before proceeding to delete the old pool.
Removing deprecated Instance poolsLink to this anchor
After verifying that workloads have been rescheduled, continue by deleting the old pool(s).
- Return to your cluster’s Pools tab and wait a few minutes to ensure all workloads have been rescheduled onto new nodes.
- Click «See more Icon» next to the deprecated pool.
- Select Delete pool.
- Confirm the deletion.
Verifying migration successLink to this anchor
-
Check your nodes:
kubectl get nodesNoteOnly nodes based on supported Instance types should now be listed.
-
Test your applications to confirm they are functioning correctly on the new nodes.
Minimize downtime by maintaining multiple replicas of key workloads and setting up Pod Disruption Budgets (PDBs).
If you require assistance during the transitioning process, please contact our Support team.