installation / basic operation questions

I had use EMS for a while and I am interested about low-latency stream. I have a customer who need low-latency live stream for their staff to monitor every level crossing of railway. However, there are thousand IP cameras and we can’t watch them all in the same time. They need the videos are shown in sub-second when they switch displayed cameras. I list two attentions of customer’s environment as the followings. 1. All IP cameras use LTE to be network environment 2. When they switch to one camera or some cameras, the live video must be displayed as soon as possible (The toleration is 1~2 seconds maximal) There are some questions that we need to be answered, 1. How many cameras can insert into a EMS？ 2. Dones EMS support HA (or Failover) ? 3. The document say EMS can record a camera streaming to a video file. Assume it record for a week which 24 hours each day. Does it only a huge video file or separate to many small video files? 4. How can I playback the recorded file? 5. If RTSP and P2P of EMS are low-latency, what’s benefit of P2P? Does it just has NAT traversal ? 6. Assume EMS is behind NAT and I turn P2P (WebRTC) on. Why can I see my camera’s streaming on “http://ers.evostream.com:5050/demo/evowrtcclient.html?stream=myStreamName&room=myRoomID” if my laptop is not behind the same NAT ? (If my laptop is behind the same NAT I can see streaming by my browser using the url.)