Assess and categorize the availability requirements for the sets of application planned for deployment on Cloud Foundation on VxRail to determine the most appropriate architecture and design.

• Determine the operational recovery and disaster recovery objectives of the application sets planned for Cloud Foundation on VxRail. Certain application sets can then be placed in VI workload domains that are configured to support these objectives.
• A consolidated architecture may be preferable to a standard architecture under these conditions:

• The planned workload can share resources with the management domain
• Minimal growth is planned
• No specialized applications are required

• A deployment with multiple availability zones may be preferable to a single site under these conditions:

• Applications targeted for the Cloud Foundation on VxRail instance cannot be impacted by a loss of service or a loss of data if a site failure occurs.
• The network latency and bandwidth requirements between the sites can be met.
• A third site is available for the stretched cluster ‘witness’, which monitors the stretched vSAN datastore.
• All of the applications planned for the Cloud Foundation on VxRail instance can run on a single site if an outage occurs.

• These conditions are suitable for a multi-region architecture:

• A longer distance between sites is desired compared to what is supported with a multi-availability architecture to support disaster recovery.
• A higher RPO and RPO is tolerable compared to what is supported with a multi-availability architecture.
• A multi-region architecture is compatible with single site and multi-site Cloud Foundation on VxRail instances to offer extended protection for applications.

• A VI workload domain at a remote site is applicable under the following conditions:

• The remote sites lack administrative capacity to manage a full Cloud Foundation on VxRail instance.
• The workload size targeted for the remote site can fit within the resource pool provided by four VxRail nodes.
• The network between the central site and remote sites meets latency and bandwidth requirements.