How it works...

When planning for highly available VMM management servers, you should first consider
where to place the VMM cluster. As per best practices, the recommendation is to install
the VMM cluster on a management cluster, preferably on some physical servers, if using converged network for your virtual network. However, if you plan to install highly available VMM management servers on the managed cluster, you need to take into consideration the following points:

• Only one highly available VMM management server is allowed per Failover Cluster.
• Despite the possibility to have a VMM management server installed on all cluster nodes, only one node can be active at a time.
• To perform a planned failover, use Failover Cluster Manager. The use of the VMM console is not supported.
• In a planned failover situation, ensure that there are no running tasks on the VMM management server, as it will fail during a failover operation and will not automatically restart after the failover operation.
• Any connection to a highly available VMM management server from the VMM console will be disconnected during a failover operation, reconnecting right after.
• The Failover Cluster must be running Windows Server 2016 in order to be supported.
• The highly available VMM management server must meet system requirements. For information about system requirements for VMM, see the Specifying the correct system requirements for a real-world scenario recipe in this chapter.
• In a highly available VMM management deployment, you will need a domain account to install and run the VMM management service. You are required to use distributed key management (DKM) to store the encryption keys in Active Directory.
• A dedicated and supported version of Microsoft SQL Server should be installed. For supported versions of SQL Server for the VMM database, see the Specifying the correct system requirements for a real-world scenario recipe.