High-availability and Disaster Recovery

For production environments, it is recommended to use an active-active high-availability approach to distribute the processing load of agent communication. This lets the agents communicate across multiple servers when many agents are communicating at once, rather than relying on the processing of a single server, providing high server availability and therefore faster response times.

The following sections describe how to configure high-availability and disaster recovery implementations:

Recommended, truly scalable method: agents connect to a single endpoint that uses round-robin DNS or an agent relay; see Connecting Agents to a Single Endpoint.
Adequate for smaller or test installations: agents connected to a series of endpoints (not recommended for large, scalable installations); see Connecting Agents to a Series of Server Endpoints.
High-availability disaster recovery: Use a hot standby server if you need to maintain high availability in the case of system failure; see Disaster Recovery with Hot Standby.
Cold standby disaster recovery: Use a cold standby server if you do not need to maintain high availability in the case of system failure; see Disaster Recovery with Cold Standby.