Description

VM Host Affinity supports running selected VM groups on different physical host groups. For example, run service group A of the active-active service in the main fault domain of the stretched cluster, and run service group B of the active-active service in the secondary fault domain of the stretched cluster to ensure that data and services are active-active at the same time.

Precautions

1. The running location must be configured in the same fault domain or automatically selected for the same VM group.
2. A running virtual machine does not support modifying the Running Location configuration.
3. After the scheduling policy is created for the running virtual machine, you need to restart the virtual machine to take effect.
4. A stretched cluster group hosts in the same fault domain into the same host group by default.

Prerequisites

None.

Steps

1. Go to the Reliability > VM Scheduling > VM Group interface, and add the virtual machines planned to run in the same location to the VM group.
2. Go to the Reliability > VM Scheduling > Node Group interface and add the nodes in the primary and secondary fault domains to the same node group.
3. Select VM-Host Affinity for Policy Type.

• If you select Must for Automation Level, VMs in VM group A will always run in the primary fault domain, and those in VM group B will always run in the secondary fault domain.

• If you select Prefer to for Automation Level, you can set a standby host group. In this case, if no host is available in the selected host group, VMs will be scheduled to hosts in the selected standby host group.