Incident detector cached data and updating service

[Moi_Nccncc]

Hello,

   How Incident detector handles his cached data when updating the service? 

   Suppose I am having a running service which has Incident detector accumulative, If I update the service dose Incident detector looses all events and starts over? 

           Example: If Truck stopped create an Incident and wait for 300 sec.  If Truck A stopped and I updated my service dose Truck A not any more in the Incident detector? 
And As general Quesition

 What is the best scenario and best practices to update a service in operational geo event server 

[GregoryChristakos]

@Moi_Nccncc - Incidents from the Incident Detector are managed in a separate cache/queue across all GeoEvent Services. The only way these incidents can be lost is by restarting the GeoEvent Server services (Windows services) or by restarting the machine where GeoEvent Server is running. Stopping and starting a GeoEvent Service, or republishing a GeoEvent Service isn't going to purge these incidents.

Given this information, you should be able to update your GeoEvent Services per the normal pattern of use. In certain situations where an Incident isn't making logical sense anymore (e.g. an Incident is reported as ongoing from before the change, and you're trying to test/confirm that your latest changes are working and so the incident should be closed), it can be advantageous to restart GeoEvent Server too.