NSX Manager Cluster Requirements

The following subsections describe cluster requirements for NSX appliances and provides recommendations for specific site deployments.

Cluster Requirements

• In a production environment, the NSX Manager (Local Manager in an NSX Federation environment) or Global Manager cluster must have three members for scaling out and for redundancy of its management and control planes.

  Each cluster member should be placed on a unique hypervisor host with three physical hypervisor hosts in total. This is required to avoid a single physical hypervisor host failure impacting the NSX control plane. It is recommended you apply anti-affinity rules to ensure that all three cluster members are running on different hosts.

  The normal production operating state is a three-node cluster of the NSX Manager (Local Manager in an NSX Federation environment) or Global Manager. However, you can add additional, temporary nodes to allow for IP address changes.

  Important: NSX Manager cluster requires a quorum to be operational. This means at least two out of three of its members must be up and running. If a quorum is lost, NSX management and control planes do not work. This results in no access to the NSX Manager's UI and API for configuration updates using the management plane as well as no access to add or vMotion VMs on NSX segments. The dynamic routing on Tier-0 remains operational.

• Starting with NSX 3.1.1, NSX supports the deployment of a single NSX Manager or Global Manager in production deployments. To recover from a failed NSX Manager, configure vSphere HA for the single NSX Manager. The time required to recover a single NSX Manager that uses backup/restore or vSphere HA might be much longer than the availability provided by a cluster of NSX Manager nodes. Similarly, for lab and proof-of-concept deployments where there are no production workloads, you can run a single NSX Manager or Global Manager to save resources. NSX Manager or Global Manager nodes can be deployed on either ESXi or KVM. However, mixed deployments of managers on both ESXi and KVM are not supported.

• If you deploy a single NSX Manager or Global Manager in a production environment, you can also enable vSphere HA on the host cluster for the manager nodes to get automatic recovery of the manager VM in case of ESXi failure. For more information on vSphere HA, see the vSphere Availability guide in the vSphere Documentation Center.
• If one of the NSX Manager node goes down, the three node NSX Manager cluster provides redundancy. However, if one of the NSX Manager node loses its disk, it goes into read only state. It might still continue to participate in the cluster if its network connectivity is Up. This case can result in the management plane and control plane to become unvailable. To resolve the issue, power off the impacted NSX Manager.

Configure Anti-affinity rules in VMware vCenter

Set up DRS Anti-Affinity rules to prevent, whenever possible, two NSX Manager VMs from running on the same host. Follow the vSphere guide to configure VM to VM affinity rules. See Create a VM-VM Affinity Rule.

VM to Host affinity rules. See Create a VM-Host Affinity Rule.

Recovery with vSphere HA

You can use vSphere HA (High Availability) with NSX to enable quick recovery if the host running the NSX Manager node fails. See Creating and Using vSphere HA Clusters in the vSphere product documentation.