Configuring high availability

You can configure high availability settings of the cluster in its card. Enter Clusters → select the cluster → High availability.

Before enabling high availability, make sure that the cluster meets the requirements and recommendations from the article Operating principle of high availability clusters.

Initial configuration

Click Enable high availability button and specify the settings:

Select which VMs should be recovered when the cluster node fails:
- All VMs in the cluster with the network storage;
- All VMs of the selected nodes → select the cluster nodes;
- VMs by parameters:
  1. Select the VMs from the list.
  2. Specify the Start priority. First, the platform recovers VMs with a higher priority value.
  3. Click Select .. VM.
If necessary, specify the Checking IP for VM. If connection to the cluster is lost, the node will check the availability of that IP address using the ping utility. If the IP address is unavailable, the node will be isolated and the process of VM relocation will start. If the IP address is available, the node will be excluded from the high availability cluster and the VMs on that node will continue to operate.
Click Save button.

Example of configuration

To improve cluster fault tolerance, you can reserve some RAM on the nodes for cluster needs. To do this, set the RAM overselling ratio for the cluster to less than 1. For example, if you set the coefficient to 0.9, 10% of the RAM will be reserved.

"VM for recovery" section

The "VM for recovery" section displays the list of VMs that should be recovered if any cluster node fails. You can perform operations with the VMs through the menu. To perform group operations, select the VM and choose the required operation from the list. For example, Restart or Change priority.

Section interface

New VMs are not automatically added to this list. If you selected the option All VMs in the cluster with the network storage or All VMs of the selected nodes when you enabled high availability, the option applies only to VMs that have already been created.

To change the list of VMs, click Change members:

Select the VMs from the list.
Specify the Start priority. First, the platform recovers VMs with a higher priority value. For example, a VM with a priority of 200 will be restored earlier than a VM with a priority of 100.
Click Select .. VM.

"Cluster nodes" section

The "Cluster nodes" section displays the list of nodes included in the high availability cluster. Nodes appear in this list automatically after being added to the cluster. You can perform operations with the nodes through the menu. To perform group operations, select the nodes and choose the required operation from the list. For example, Start script.

This section displays the statuses of the ha-agent services on these nodes, not the status of the nodes themselves. To find out the status of a node in the cluster, go to the Nodes section.

Section interface

"Settings" section

In the "Settings" section, you can change the preset settings or Disable high availability.

Cleaning up artifacts on a cluster node

If the high availability disabling failed, there may be artifacts left on the cluster nodes — VM files that were not deleted after the relocation. In this case, the problem nodes display the sub-status "Problem disabling HA". To remove artifacts, click this sub-status → Clear HA artifacts on the node button.

Useful tips