Multi-machine services follow the same general life cycle as individual machines, from requesting and provisioning through managing and decommissioning.
A multi-machine blueprint contains references to blueprints for the component machines. For each component blueprint, it specifies the minimum and maximum number of machines for the multi-machine service.
A multi-machine blueprint can specify scripts, or workflows, to run during the provisioning process or when powering the multi-machine service on or off. Distributed Execution Manager worker services or agents, not the component machines, run the scripts.
When users request a multi-machine service, they can specify the following settings:
How many component machines to provision, based on limits specified in the multi-machine blueprint
Specifications for the component machines, such as CPU, memory, and storage, based on the individual component blueprints
General settings, such as lease duration and custom properties, to be applied to component machines in the multi-machine service
Before provisioning the multi-machine service, vRealize Automation allocates resources for all of the component machines. If the multi-machine service causes a reservation to become over-allocated, its provisioning fails. After the resources are allocated, the component machines are provisioned and powered on.
After the multi-machine service is provisioned, the machine owner can perform machine menu tasks on the multi-machine service as a whole, such as powering the multi-machine service off and on, or destroying the multi-machine service and its component machines.
If the multi-machine blueprint allows for a varying number of machines for any component type, the machine owner can add or delete machines from the multi-machine service after it is provisioned.
The machine owner can view the components that make up a multi-machine service and manage them as a group or individually. Most machine operations are available for individual component machines, except for changing the owner or lease.
These operations are inclusive to the multi-machine service and modify the group as a whole. Conversely, some actions are inclusive to the individual components, such as suspending, redeploying, and connecting using Microsoft Remote Desktop Protocol or VMware Remote Console.