2 of 2 people found this helpful
We have some news on this. We had opened a case with BMC Support and together we discovered that this is a known error. If for any reason an event cannot be processed by the RE it remains in the queue for a second processing. After the second processing, if it is successful or not, the event should be removed from the queue. However, this is not happening. Everytime when a RE starts, it re-processes the events from the queue, what leads to long delays until new events can be processed. According to the BMC Support, this will be fixed in the next release.
The queue is in fact a table in the TSSR RE database called EVENT_MSGS. It is possible to manually delete events from this table, to prevent long delays after restarting the RE.