Automated Monitoring, Alerting, and Recovery for Stalled TRIRIGA Agents

When a TRIRIGA agent stops functioning, it may still appear as RUNNING in the Administrator Console. There is currently no automated notification to alert administrators when this happens. As a result, these issues often go unnoticed until users report problems with the system. In many cases, the issues identified in the server.log can be resolved by simply restarting the affected agent.

For example, we have experienced recurring issues with the Future Event Workflow agent. When abnormal behavior is observed, such as workflows not processing, we investigate the server.log and often find errors such as deadlocks. Before resorting to a full application restart, we typically attempt to restart the affected agents to restore normal processing.

The proposed enhancement would introduce automated monitoring, notifications, and self-healing capabilities for TRIRIGA agents:

• Automated notifications: Alert administrators when an agent stops functioning or becomes unresponsive, even if it still appears as running.
• Automatic recovery: Detect stalled agents and automatically restart or recover them when possible, without requiring a full application restart.
• Improved agent health monitoring: Provide visibility into the actual operational status of agents, rather than relying solely on their running status in the Administrator Console.