Performing In-place Upgrade

Description

• An in-place upgrade updates the Kubernetes components on cluster nodes, without changing their OS version.
• Data plane nodes are upgraded in batches. By default, they are prioritized based on their CPU, memory, and PodDisruptionBudgets (PDBs). You can also set the priorities according to your service requirements.

Precautions

• During the cluster upgrade, the system will automatically upgrade add-ons to a version compatible with the target cluster version. Do not uninstall or reinstall add-ons during the cluster upgrade.
• Before the upgrade, ensure that all add-ons are running. If an add-on fails to be upgraded, rectify the fault and try again.
• During the upgrade, CCE checks the add-on running status. Some add-ons (such as CoreDNS) require at least two nodes to run normally. In this case, at least two nodes must be available for the upgrade.

For more information, see Before You Start.

Procedure

The cluster upgrade goes through check, backup, configuration and upgrade, and verification.

1. Log in to the CCE console and click the cluster name to access the cluster console.
2. In the navigation pane, choose Cluster Upgrade.
3. Perform the pre-upgrade check. Click Start Check and confirm the check. If there are abnormal or risky items in the cluster, handle the exceptions based on the check results displayed on the page and check again.
   • Exceptions: View the solution displayed on the page, handle the exceptions and check again.
   • Risk Items: may affect the cluster upgrade. Check the risk description and see whether you may be impacted. If no risk exists, click OK next to the risk item to manually skip this risk item and check again.

   After the check is passed, click Next: Back Up.

4. (Optional) Manually back up the cluster data. Data is backed up during the upgrade following a default policy. You can click Back Up to manually back up data. If you do not need to manually back up data, click Next: Configure & Upgrade.

   Manual backup will back up the entire master node. The backup process uses the Cloud Backup and Recovery (CBR) service and takes about 20 minutes. If there are many cloud backup tasks at the current site, the backup may take longer. The cluster cannot be upgraded during the backup.

5. Configure the upgrade parameters.
   • Add-on Upgrade Configuration: Add-ons that have been installed in your cluster are listed. During the cluster upgrade, the system automatically upgrades the add-ons to be compatible with the target cluster version. You can click Set to re-define the add-on parameters.
     

     If a red dot is displayed on the right of an add-on, the add-on is incompatible with the target cluster version. During the upgrade, the add-on will be uninstalled and then re-installed. Ensure that the add-on parameters are correctly configured.

   • Node Upgrade Configuration: You can set the maximum number of nodes to be upgraded in a batch.
   • Node Priority: You can set priorities for nodes to be upgraded. If you do not set this parameter, the system will determine the nodes to upgrade in batches based on specific conditions. Before setting the node upgrade priority, select a node pool. Nodes and node pools will be upgraded according to the priorities you specify.
     • Add Upgrade Priority: Add upgrade priorities for node pools.
     • Add Node Priority: After adding a node pool priority, you can set the upgrade sequence of nodes in the node pool. The system upgrades nodes in the sequence you specify. If you skip this setting, the system upgrades nodes based on the default policy.

6. After the configuration is complete, click Upgrade and confirm the upgrade. The cluster starts to be upgraded. You can view the process in the lower part of the page.

   During the upgrade, you can click Suspend on the right to suspend the cluster upgrade. To continue the upgrade, click Continue. When the progress bar reaches 100%, the cluster upgrade is complete.

   

   If an upgrade failure message is displayed during the cluster upgrade, rectify the fault as prompted and try again.

7. After the upgrade is complete, click Next: Verify. Verify the upgrade based on the displayed check items. After confirming that all check items are normal, click Complete and confirm that the post-upgrade check is complete.

   You can verify the cluster Kubernetes version on the Clusters page.