From each replication topology, pick one server (this could be master or slave) and let orchestrator know which hostname & port this server listens on. When considering automated recovery, downtimed servers are ignored. Furthermore, it will place a recovery lock on the instance.
The following assumes you will be using the same machine for both the orchestrator binary and the MySQL backend. If not, replace 127.0.0.1 with appropriate host name. Topology changes require placing locks on the minimal set of affected instances, so as to avoid an incident of two uncoordinated operations on a smae instance (leading to possible chaos). Locks are placed in the backend database, and so multiple orchestrator instances are safe. This will enable SSL via the web interface (and API) so that communications are encrypted, like a normal HTTPS web page.