We should add more logs: - reason to failover coordinator leader change - additional info about instances -- not only the uuids, but uris