You can deploy the Remote Management Server (RMS) in four modes. In most cases the Two Instances (Active – Passive Servers) mode is the preferred method.
The following is a list of the three available modes and a description of each.
Single RMS processes all requests. This mode is the simplest configuration.
Two Instances (Active – Passive Servers)
Both servers (an active primary server and a passive secondary server) run behind a load balancer. The load balancer periodically checks the health of the primary and secondary servers. If the primary server is deemed to be down, the load balancer switches all the requests to the secondary server until the primary is back online.
This method ensures that high availability is maintained. This method is suggested when there is a low number of concurrent RM connections or when the load balancer does not support a persistent session.
Horizontal Scaling with Multiple Instances (Active – Active Servers)
RMS instances using session persistence on a load balancer is the optimal solution for maintaining growth and performance. Multiple RM servers run behind a load balancer. Unlike the “active-passive” deployment above, this method is preferred when network traffic from balancing connections between multiple RM sessions is more important than just high availability.