Kubernetes Engine support for node cycling extended to include reboot, and boot volume replacement

OCI Kubernetes Engine now enables you to resolve issues and perform maintenance tasks on the compute instances hosting worker nodes, by extending support for node cycling to include reboot, and boot volume replacement.

When you cycle worker nodes, you can now:

• Reboot specific managed nodes and self-managed nodes (using the OCI OKE API, or directly through the Kubernetes API).
• Replace boot volumes of specific managed nodes and self-managed nodes (using the OCI OKE API, or directly through the Kubernetes API), or replace boot volumes of all the managed nodes in a managed node pool (using the OCI OKE API).

These Kubernetes Engine features leverage Kubernetes API cordon and drain functionality, enabling you to maintain system availability during reboot and boot volume replacement operations.

When you cycle worker nodes in the Console, you are also given options to terminate and replace specific managed nodes (the existing node deletion feature), or to terminate and replace all the worker nodes in a node pool (the existing node cycling feature).

As well as cycling worker nodes to resolve issues and perform maintenance tasks, you can also cycle all the managed nodes in a managed node pool (replacing boot volumes, or terminating and replacing nodes, as appropriate) when you want to:

• Update the properties of all the managed nodes.
• Upgrade managed nodes to a newer version of Kubernetes.

Previously, if you wanted to update the Kubernetes version, image, or public SSH key properties of nodes in a managed node pool, you had to terminate and replace the nodes. With the newly extended node cycling feature, you can now update the Kubernetes version, image, and public SSH key simply by replacing managed node boot volumes. Replacing boot volumes of managed nodes is significantly faster than terminating and replacing the nodes, especially for nodes with bare metal shapes.

For more information, see:

• Rebooting Worker Nodes
• Replacing Boot Volumes of Worker Nodes
• Terminating and Replacing Worker Nodes
• Performing an In-Place Worker Node Update by Cycling Nodes in an Existing Node Pool
• Performing an In-Place Managed Node Kubernetes Upgrade by Cycling Nodes in an Existing Node Pool