A configuration was improperly applied that caused a route leak into our internal routing protocols. This caused several connection issues and we ended up restarting our edge routers to fix the issue. This resulted in an outage for approximately 20 min while they rebooted.