Performing Replace/Rolling Upgrade (v1.13 and Earlier)

Scenario

You can upgrade your clusters to a newer Kubernetes version on the CCE console.

Before the upgrade, learn about the target version to which each CCE cluster can be upgraded in what ways, and the upgrade impacts. For details, see Overview and Before You Start.

Precautions

• If the coredns add-on needs to be upgraded during the cluster upgrade, ensure that the number of nodes is greater than or equal to the number of coredns instances and all coredns instances are running. Otherwise, the upgrade will fail. Before upgrading a cluster of v1.11 or v1.13, you need to upgrade the coredns add-on to the latest version available for the cluster.
• When a cluster of v1.11 or earlier is upgraded to v1.13, the impacts on the cluster are as follows:
  • All cluster nodes will be restarted as their OSs are upgraded, which affects application running.
  • The cluster signing certificate mechanism is changed. As a result, the original cluster certificate becomes invalid. You need to obtain the certificate or kubeconfig file again after the cluster is upgraded.
• During the upgrade from one release of v1.13 to a later release of v1.13, applications in the cluster are interrupted for a short period of time only during the upgrade of network components.
• During the upgrade from Kubernetes 1.9 to 1.11, the kube-dns of the cluster will be uninstalled and replaced with CoreDNS, which may cause loss of the cascading DNS configuration in the kube-dns or temporary interruption of the DNS service. Back up the DNS address configured in the kube-dns so you can configure the domain name in the CoreDNS again when domain name resolution is abnormal.

Procedure

1. Log in to the CCE console. In the navigation pane, choose Resource Management > Clusters. In the cluster list, check the cluster version.
2. Click More for the cluster you want to upgrade, and select Upgrade from the drop-down menu.
   

   • If your cluster version is up-to-date, the Upgrade button is grayed out.
   • If the cluster status is Unavailable, the upgrade flag in the upper right corner of the cluster card view will be grayed out. Check the cluster status by referring to Before You Start.

3. In the displayed Pre-upgrade Check dialog box, click Check Now.
4. The pre-upgrade check starts. While the pre-upgrade check is in progress, the cluster status will change to Pre-checking and new nodes/applications will not be able to be deployed on the cluster. However, existing nodes and applications will not be affected. It takes 3 to 5 minutes to complete the pre-upgrade check.
5. When the status of the pre-upgrade check is Completed, click Upgrade.
6. On the cluster upgrade page, review or configure basic information by referring to Table 1.

   Table 1 Basic information

   Parameter	Description
   Cluster Name	Review the name of the cluster to be upgraded.
   Current Version	Review the version of the cluster to be upgraded.
   Target Version	Review the target version after the upgrade.
   Node Upgrade Policy	Replace (replace upgrade): Worker nodes will be reset. Their OSs will be reinstalled, and data on the system and data disks will be cleared. Exercise caution when performing this operation. NOTE: The lifecycle management function of the nodes and workloads in the cluster is unavailable. APIs cannot be called temporarily. Running workloads will be interrupted because nodes are reset during the upgrade. Data in the system and data disks on the worker nodes will be cleared. Back up important data before resetting the nodes. Data disks without LVM mounted to worker nodes need to be mounted again after the upgrade, and data on the disks will not be lost during the upgrade. The EVS disk quota must be greater than 0. The container IP addresses change, but the communication between containers is not affected. Custom labels on the worker nodes will be cleared. It takes about 20 minutes to upgrade a master node and about 30 to 120 minutes to upgrade worker nodes (about 3 minutes for each worker node), depending on the number of worker nodes and upgrade batches.
   Login Mode	Key pair: Select the key pair used to log in to the node. You can select a shared key. A key pair is used for identity authentication when you remotely log in to a node. If no key pair is available, click Create a key pair.

7. Click Next. In the dialog box displayed, click OK.
8. Upgrade add-ons. If an add-on needs to be upgraded, a red dot is displayed. Click the Upgrade button in the lower left corner of the add-on card view. After the upgrade is complete, click Upgrade in the lower right corner of the page.
   

   • Master nodes will be upgraded first, and then the worker nodes will be upgraded concurrently. If there are a large number of worker nodes, they will be upgraded in different batches.
   • Select a proper time window for the upgrade to reduce impacts on services.
   • Clicking OK will start the upgrade immediately, and the upgrade cannot be canceled. Do not shut down or restart nodes during the upgrade.

9. In the displayed Upgrade dialog box, read the information and click OK. Note that the cluster cannot be rolled back after the upgrade.
10. Back to the cluster list, you can see that the cluster status is Upgrading. Wait until the upgrade is completed.

    After the upgrade is successful, you can view the cluster status and version on the cluster list or cluster details page.