Node draining in Rancher-provisioned RKE2 and K3s clusters upon Rancher version upgrade
Article Number: 000022311
Environment
- A Rancher-provisioned RKE2 or K3s cluster, with Drain Nodes enabled in the Upgrade Strategy.
- A Rancher version upgrade.
Situation
When performing a Rancher version upgrade, users may observe that Rancher-provisioned RKE2 or K3s clusters, with Drain Nodes enabled in the Upgrade Strategy, are updated after the Rancher upgrade and node drains are performed as a result.
Cause
As noted in the Rancher documentation on enabling node drains:
During an upgrade, nodes may be drained even when no user-visible YAML changes are present. This can occur if non-dynamic configuration files are updated or if a new system-agent-installer image is introduced. In such cases, Rancher generates a new upgrade plan, resulting in a new plan hash. When Upgrade Strategy is set to Drain nodes, this plan change can trigger node draining.
Resolution
If users wish to avoid the possibility of nodes drains - which will cause temporary workload unavailability due to rescheduling - when upgrading Rancher, set Drain Nodes to No in the Upgrade Strategy of the clusters, before the Rancher upgrade:
- Navigate to Cluster Management in the Rancher UI and click Edit Config for the relevant RKE2/K3s cluster.
- Under Cluster Configuration select Upgrade Strategy.
- Set the Drain Nodes option to No for both Control Plane and Worker Nodes.
- Click Save to apply the changes.
After the Rancher upgrade, you can revert this change, to set Drain Nodes back to Yes.