Orion servers can be deployed standalone or with multiple instances across a distributed network. Below, Figure 1 illustrates a configuration in which a single Orion server monitors a large corporate network, as well as a smaller branch network. This straightforward deployment approach uses a single SQL database. Monitoring scalability depends on database performance, which typically limits the number of monitored network elements to between 50,000 and 100,000, depending on the polling frequency and other factors. This deployment configuration is ideal when an organization and its IT staff are geographically centralized.
When multiple Orion servers are used for management, Orion EOC can be leveraged as a command center console to aggregate the multiple instances, as shown in Figure 2. This configuration is ideal for enterprise-class organizations with geographically distributed networks or for a MSP (managed service provider) managing multiple networks.
In this deployment, each Orion server monitors its local network and feeds aggregate information back to EOC. The corporate office has an Orion server with a polling engine to extend its monitoring scalability. The East Coast and West Coast networks each use a single Orion server to monitor their local networks and minimize international WAN traffic back to the primary corporate network.
Orion EOC securely collects Orion server data directly from each of the regional SQL databases, including the local corporate office. WAN performance is not impacted because Orion servers poll network devices locally and EOC only periodically pulls updates from each Orion server database. This WAN-optimized architecture ensures that WAN traffic is minimized and that, even if the WAN link temporarily goes down, regional Orion servers will continue polling without disruption. Once the WAN link is restored, EOC automatically reconnects to the Orion servers, ensuring you never lose important information about network health.