Every time a core VOIP server goes down for a few seconds, every single remote site sends out an alarm, and therefore generates an incident when they go into standalone survivable mode, even if they reliquish control back to the core server 30 seconds later. And every one of those alarms has to...