Simplify failover in HA

The manual failover of a FDB HA cluster is currently pretty complicated and error prone without additional tooling since the operator has to gather the complete database configuration. Then change the priority of the main and remote DC. Finally the operator can submit the configuration change (and hopefully the operator didn't changed any other settings). It would be great to simplify this process for an operator to be able to directly call e.g. `fdbcli fail-over $dc` or `fdbcli mark-as-failed $dc` (these are obviously only ideas 😄).

Start with the existing fdbcli configuration and failover-related entry points, then review how operators currently change database priorities and submit configuration updates. Define the command semantics and safety requirements for directly failing over or marking a data center failed; done requires an agreed design and an implemented, tested workflow.

Written by the indexing model from the issue text.