When a Server starts up or is rebooted the Symantec Scan Engine 5.2 service fails to start in a timely fashion, and the Scan Engine service has to be manually started again after the Server is up.
Typically we see this issue occurs in a heavily loaded environment. Such as a Server that is running several CPU intensive processes, with the Server itself being hosted by a heavily loaded vmware server.
One solution to help with this is to try increasing the Service Control Manager timeout period. The default SCM timeout period is 30 seconds, and if a service does not report to the SCM within this timeout period the SCM will terminate that process. Below is a link with details on how to change the Service Control Manager timeout period,