You can deploy the Incident Manager role on a standalone machine and configure failover on a separate standalone machine.
The procedure to set up failover clusters for the Incident Manager role follows the same process as configuring role failover in Security Center. For more information, see Role failover and Setting up role failover.
Benefits: Incident Manager and failover configuration on separate machines
- Incident Manager resources are not shared
- Machine resources are not shared even in failover mode and therefore do not impact the Incident Manager performance or availability.
- Fault resilience for the Incident Manager role
- If the machine running the Incident Manager goes offline, the system launches another instance of the Incident Manager on the machine with the other services after a slight delay thereby resuming operations.
- Minimal downtime for rolling updates
- Updates for your Windows system are available with minimal downtime for Incident Manager role.
Drawbacks: Incident Manager and failover configuration on separate machines
- Higher complexity in deployment
- You must install the Security Center server in two additional machines, move the Incident Manager role to one and configure Incident Manager failover in the other.
- Medium maintenance effort
- There are three machines in this deployment architecture to monitor and maintain.
- Higher latency in communication between components
- There might be a slight delay in communications between the Incident Manager role and the other services.
- No fault resilience or fault tolerance for other components
- If any of the other components go offline, the lack of support for failover or high availability for them leads to data loss and service unavailability.
Quality of service
Using this deployment architecture implies that service quality relies more on your system infrastructure. The quality of service would be directly dependent on the quality and reliability of your network infrastructures.