When you create an ESX agency, you must define the scope of the agency. You define the scope of an ESX agent in terms of compute resources, namely standalone hosts or clusters.
For example, ESX agents can run on standalone ESXi hosts or on clusters of ESXi hosts.
A solution must obtain from vCenter Server the set of compute resource instances on which to deploy ESX agents. For example, you can implement a query in your solution that obtains a list of all the hosts that are running in a vCenter Server instance. Users can then select the hosts on which to deploy ESX agents.
You set the initial ESX agency scope in the ESX agency configuration. As the solution runs, the scope of the ESX agency can change. For example, your solution runs on compute resources that users select from your solution Configuration page. As users select different compute resources on which to run the solution, the scope of the ESX agency changes as it adds the MoRefs of the selected compute resources to the scope.