Active-Passive Architecture
The following diagram shows the structure of an FME Server system properly configured for active-passive failover.
An Active-Passive system provides separate machines for the following components, which should be configured for fault tolerance:
Additionally, two separate machines - one primary and one failover - are provided on which to install the FME Server Core and the FME Engines. Fault-tolerance is built in to these components.
If the machine that houses the primary FME Server Core and Engines (Machine A) fails, the Web Application Server connects to the secondary FME Server Core and Engines on Machine B. Because the failover Core is now the "active" core, clients subsequently connect to the failover Core so that processing continues. When a heartbeat failure is detected, this failover Core takes over jobs and schedules.
When machine A is restored, the Core on Machine B remains the active core, while the Core on Machine A is now online as the failover core.
Keep in mind the following when implementing an active-passive system:
-
After the active Core fails, it takes three minutes for processing to resume on the failover Core. Any Schedules that would have triggered jobs during this three-minute window will not occur.
- Any jobs currently running when the active Core fails restart automatically on the failover Core.
- Clients of Notification Service publishers —including Amazon SNS, UDP, Email (IMAP), JMS, and WebSocket—do not failover. These clients must be manually reconfigured to connect to the new active core. Alternatively, restore to the original active core.
Installing an Active-Passive System
See Also