Start, Stop and Restart Platform Service

The example service's core process can be started, stopped, and restarted through its service manager on each application server, or fleet-wide through the deployment pipeline.

Restarting a single server's service is appropriate for isolated issues such as a memory leak on one instance. Restarting the entire fleet should go through the deployment pipeline instead of being done server by server, so that the restart happens gradually and traffic is drained from each server before it restarts.

Stopping the service entirely is only done for planned maintenance, and always with advance notice on the public status page.