What happens if the Monitoring Agent service stops?
No impact on Mule runtimes—monitoring is read-only. Agent crash/restart doesn't affect monitored applications (agent only reads JMX data, doesn't control runtimes unless Remote Action sent). Monitoring gap: During agent downtime, status checks pause—no alerts sent if apps stop during this window. When agent restarts, resumes monitoring immediately, catches up on current state (doesn't replay history). High-availability: Deploy 2 agents monitoring same Mule instances—if Agent 1 stops, Agent 2 continues alerting (configure alert de-duplication to prevent double-notifications).