Clusters

Nova Node Clusters provide the user or administrator with the ability to manage a 2 node cluster configuration in active / standby mode. This enables risk mitigation through high availability.

Prerequisites

The following are required in order to set up clustering:

  • 2 or more Nodes (Virtual or Physical Machines)
    • It is recommended to run Nova as virtual machine therefore we recommend using our validated Nova virtual machine images. Download links for each supported variation are available on the installation page when adding a Nova Node.
    • Be sure to also follow the post install steps to ensure the Nova Node is connected when viewing the (Nova Nodes pageopen in new window)
  • A paid Nova subscriptionopen in new window
  • Controlled network space for the Virtual IP addresses. This is required for the cluster communications.
  • Direct on-host network control, meaning you cannot use a container at all. Use either the VM images, Cloud deployment or Ubuntu installation method. If you are running containers there are multiple options for high availability with Nova for Kubernetes, or, you can use Destinations.

Paid Feature

Nova Node Clusters are only available for paid subscriptions. Cloud providers do not allow Virtual IPs (VIPS) to be allocated, for this reason you should always use Destinations or an Autoscaler for highly available cloud deployments.

Screenshot

Create Two Nodes

  1. Log into Novaopen in new window
  2. Navigate to the nodes page
  3. Name and create a Node
  4. Deploy the Nova software to your host
    • Follow the instructions described when adding the node
  5. Verify that the node connected
    • Return to the Nodes page
    • Confirm that the status is set to online
  6. Repeat steps 3 to 5 for all the remaining nodes

Create a Cluster

  1. Navigate to the Node Clusters Page
  2. Create a Cluster
  3. Name the Cluster
  4. Set the Primary Node, Backup Node and Virtual IP to be used
  5. Verify that the cluster is up and running

You are able to create a single Node cluster to assist with VIP management or for situations where the second Node is not yet available.