Stopping and starting Kubernetes cluster

To stop the servers or virtual machines running Kubernetes (for example, to take a VMWare snapshot), follow these instructions.

Stopping the Kubernetes cluster

Procedure

To stop the cluster:

  1. As the root user, enter the following command to stop the Kubernetes worker nodes:
    Note: If running in VMWare vSphere, use Shutdown Guest OS. 
    shutdown -h now
  2. Stop all worker nodes, simultaneously or individually.
  3. After all the worker nodes are shut down, shut down the Kubernetes master node.
    Note: If the NFS server is on a different host than the Kubernetes master, you can shut down the Kubernetes master when you shut down the worker nodes.
  4. Stop the NFS server next. By default, this is located on the Kubernetes master node and will be shutdown when the Kubernetes master node host or virtual machine is powered off.
  5. Stop the server or virtual machine running the Docker registry last. The Docker registry is normally running on the Kubernetes Master node and will get stopped when the master node is powered off.

Starting the Kubernetes cluster

Procedure

To restart the cluster:

  1. Start the server or virtual machine that is running the Docker registry first. This will automatically start the Docker registry. The Docker registry is normally running on the Kubernetes Master node and will get started when Master node is started.
  2. Start the NFS server and wait two minutes after the operating system has started. The NFS server is normally on the Kubernetes Master Node.
  3. Start all worker nodes either simultaneously or individually. If the NFS server is on a dedicated host (not the Kubernetes master node), start the Kubernetes master at the same time you start the worker nodes.