NAT keepalive drops OCPP websocket

How it presents

Symptom

NAT keepalive drops OCPP websocket — sessions work briefly after reboot then Soft-offline until the next reconnect. protocol field note: keep evidence scrubbed and prove the ranked cause before module RMA on NET-C03.

Root causes (by observed frequency)

1. #1 Firewall/NAT UDP/TCP idle timeout shorter than Heartbeat

   Classic pad Wi-Fi/LTE NAT

2. #2 Heartbeat interval too long for site NAT
3. #3 Carrier CGNAT aggressiveness on cellular
4. #4 Stateful firewall depleting under multi-stall reconnect storms

Diagnostic procedure

1. 1. Plot Soft-offline cadence vs Heartbeat interval

   Expected: Idle-timeout pattern

   Tools: CSMS

2. 2. Shorten Heartbeat or enable OEM keepalive per guidance

   Expected: Stays online

   Tools: HMI/CSMS

3. 3. Ask IT to raise NAT idle timers on EVSE VLAN

   Expected: Timer > Heartbeat

   Tools: IT

4. 4. On cellular, test alternate APN/carrier path if CGNAT suspected

   Tools: modem

5. 5. Confirm ≥1 hour stable Heartbeats after change

   Tools: CSMS

The fix

Websocket drops on a timer are NAT/keepalive — adjust timers before hardware RMA.

1. Tune Heartbeat/keepalive to beat the NAT idle timer. Document readings and scrubbed photos for NET-C03 before closing the ticket.
2. Raise firewall NAT timers for the EVSE segment. Document readings and scrubbed photos for NET-C03 before closing the ticket.
3. Avoid reboot loops that only temporarily reopen the socket.
4. Document the working Heartbeat interval on site. Document readings and scrubbed photos for NET-C03 before closing the ticket.
5. Escalate modem/carrier only after timer math is fixed. Document readings and scrubbed photos for NET-C03 before closing the ticket.