Watchdog - Relay - Add maintenance mode

[fe51]

We must avoid a situation where, during manual maintenance, the camera API is restarted/down and the Pi Zero detects the service as unreachable. In such a case, the Pi Zero could interpret this as a failure and trigger a reboot of the Pi, which would interrupt the ongoing maintenance operation.

To prevent this behavior, the following mechanism should be implemented:

• Maintenance flag script on the Pi
  A script running on the Pi will send or write a timestamp to the Pi Zero indicating that a maintenance period has started.

• Maintenance window validation on the Pi Zero
  Before performing any service check or triggering a reboot, the Pi Zero must verify whether the system is currently within a maintenance window.
  The maintenance window is defined as:

maintenance_timestamp + 4 hours

If the current time falls within this window, the Pi Zero must not trigger a reboot.

• Integration with automated maintenance
  This script must also be callable during automated maintenance procedures (e.g., through Ansible) so that the maintenance window is automatically set before maintenance tasks are executed.