Best Practices: Process Management and Maintenance Consideration

• Handle uncaught exception - this capability enables a process with uncaught (unhandled) exception to be suspended instead of faulting. You can then either manually recover the process or use an alert service to handle the situation automatically.
• Set logging - there are a number of logging levels provided in the ActiveVOS runtime environment. A process deployment log provides a trace of process deployment activity. Process persistence provides a detailed description of the state of a process instance. The process log provides a chronicle of all the activities of a process instance. All logs when turned on are preserved in the database. Please note that enabling logging will result in additional overhead at runtime.
• Use ActiveVOS engine monitoring - key performance indicators such as faulted process count, count of discarded inbound messages can be monitored using the engine monitoring service. Consider enabling individual indicators and assigning a BPEL process to handle monitoring events.

Monitoring and Notifications
	Handling Suspend on Uncaught Faults using the ActiveVOS Alert Service Create, deploy and register a BPEL Process that uses the Identity and Email Services in the ActiveVOS Server to respond to alerts generated by the Alert Service
	Monitoring Engine Properties and Handling Abnormalities using the ActiveVOS Monitoring Alert Service The ActiveVOS Server provides engine monitoring, and if desired, the option to trigger an alert to notify an administrator when a monitored property has a warning or error condition. Create, deploy and register a Monitoring Alert Service BPEL Process