Register Cluster
Use register when the target third-party cluster should initiate the connection to the platform through a reverse proxy tunnel.
TOC
PrerequisitesImportant NotesRegister ClusterView Registration CommandFAQHow to Resolve Distributed Storage Deployment Failure When the Runtime Component of the Connected Cluster is Containerd?Prerequisites
-
Depending on the target environment, the Kubernetes version and component parameters in the target cluster must meet the version and parameter requirements.
-
The image registry must support HTTPS access and provide a valid TLS certificate authenticated by a public certification authority. If this requirement cannot be met, see How to Trust Insecure Image Registries?
Note: The Public Registry provided by the platform on the public network already meets HTTPS access requirements. You only need to verify whether the Platform Default and Private Registry support HTTPS access.
-
If the runtime component of the cluster to be connected is Containerd, you need to modify the Containerd configuration before connecting the cluster to ensure successful deployment of distributed storage.
Important Notes
Network interface traffic monitoring recognizes network interfaces with names matching eth\.\|en\.\|wl\.*\|ww\.* by default. If the target cluster uses other naming conventions, see Collecting Network Data from Custom Named Network Cards and update the corresponding resources after registration.
Register Cluster
-
In the left navigation bar, click Clusters > Clusters.
-
Click Managed Clusters > Register Cluster.
-
Configure the registry parameters for storing platform component images required by the registered cluster according to the following instructions.
-
Click Create, obtain the registration command on the Registration Command page and run the command in the cluster to be registered.
Note: The registration command is valid for 24 hours. Obtain a new command after it expires.
View Registration Command
Find the cluster waiting for registration in the cluster list and click View Registration Command. Run the registration command before it expires.
FAQ
How to Resolve Distributed Storage Deployment Failure When the Runtime Component of the Connected Cluster is Containerd?
When the runtime component of the connected cluster is Containerd, distributed storage deployment will fail. To resolve this issue, you need to manually modify the Containerd configuration information on all nodes of the cluster and restart Containerd.
Note: If you modify the Containerd configuration by following the steps below before deploying distributed storage, you do not need to execute step four.
-
Log in to the cluster node and edit the
/etc/systemd/system/containerd.servicefile, changing theLimitNOFILEparameter value to1048576. -
Execute the command
systemctl daemon-reloadto reload the configuration. -
Execute the command
systemctl restart containerdto restart Containerd. -
Execute the command
kubectl delete pod --all -n rook-cephon the cluster control node to restart all Pods in the rook-ceph namespace to make the configuration effective.