Verify Agent Connectivity
Goal: Confirm the robot is enrolled, the agent service is running, and telemetry is reaching the cloud (the robot shows online in the web app).
Before you begin
- The bootstrap command has been run on the robot — this installs the agent and enrolls it in one step.
- Terminal access with
sudoon the robot.
Steps
- Confirm enrollment. Check the device certificate and
robot_idexist:
bash
sudo ls -l /var/lib/rdyn/robot/device.crt
sudo grep robot_id /var/lib/rdyn/robot/robot.toml
- Check the service is running. Use the agent's systemd unit name:
bash
sudo systemctl status rover-agent
It should show active (running). If it is not enabled on boot:
bash
sudo systemctl enable --now rover-agent
- Watch the logs for startup, enrollment, and a successful cloud connection:
bash
sudo journalctl -u rover-agent -f
Look for the agent loading its config, establishing the mutual-TLS Zenoh link, and beginning to publish telemetry. Errors about a missing cert/key mean enrollment did not complete.
- Restart after config changes so they take effect:
bash
sudo systemctl restart rover-agent
- Confirm the robot is online in the Rover Nexus web app. Once telemetry uplink is flowing, the robot appears online there.
!!! note Telemetry does not flow until your robot software is publishing on the local Zenoh topics (see Map Telemetry) or the ROS 2 bridge is running. Before you have completed those steps, "no telemetry yet" is expected even when the agent is enrolled and connected.
Troubleshooting
- Not online / no uplink: verify network egress to the cloud and that the device cert and key are present.
- Service keeps restarting: read
journalctlfor the failing step (config parse, TLS, or Zenoh connect). - No telemetry: confirm robot software is publishing on the local Zenoh topics the agent ingests (see Map Telemetry).
Next steps