Create runbook-style documentation for a failover with two sites, multi-node Geo, omnibus-managed database

Provide a clear, linear process for a failover with two sites, multi-node Geo, omnibus-managed database.

• Preflight checks
• Make primary unavailable (maintenance mode)
• Promote secondary site to primary
• Optional: Demote primary to secondary and add to new primary

This may include information that is repeated in other sections!