In the case of a failure or some event described below, the rule engine may suddenly stop operating, that is, stop reading events from the Event Channel and processing them.
Internal exceptions. Such as division by 0. In such cases, the error is considered serious enough to stop processing the current event. An error is logged (in bpmevents.log)
Communication failures. For example, the database connection may be broken, and the retry policy—as specified in the configuration file—has been unsuccessful. Or, calls to the BP Server process interface are failing (as well as the retry policy for them). In such case, an error is also logged. Note that in this case, as in any failure involving a Java invocation, the Java exceptions may be caught at the BPM Events language level. So a rule could catch such failures, and proceed further. If no exception is caught, then this is a case where the server stops.
In such cases, the engine stops processing and the event reader remains in a suspended mode. However, the status of the server, when queried through the BPM Events Admin utility (described below) or through the server API, shows "STARTED", instead of "STOPPED" or "SUSPENDED".
In order to fix the problem, the you can proceed with one of the following scenarios.
Internal exceptions. Such as division by 0. In such cases, the error is considered serious enough to stop processing the current event. An error is logged (in bpmevents.log)
