Enable Zenoh shared-memory transport

Understand the shared-memory experiment

By default, processes on the same machine exchange data over TCP loopback. Zenoh shared-memory transport places large messages in /dev/shm and avoids the network stack. Applications don’t need code changes or loaned buffers, and Zenoh falls back to TCP if shared memory is unavailable.

You first measure the default TCP-loopback latency. You then enable shared memory in both Zenoh configuration files, restart the relevant processes, and repeat the measurement.

Measure the TCP-loopback baseline

Stop Navigation2 in Terminal 3. The latency measurement needs wall-clock timestamps, and Navigation2 doesn’t operate in this mode.

Stop the current simulation. Restart it in Terminal 2 with wall-clock time and without the Gazebo viewer:

    

        
        
just rox_simu use_wall_time:=True no_gui

    

Open a new terminal, source the environment, and measure the point-cloud latency:

    

        
        
source ~/workshop_env.bash
just cam_latency

    

On the supplied reference system, the point-cloud latency is around 9–10 ms:

    

        
        Mean : 9.40 ms | Std : 0.82 ms | Min : 7.91 ms | Max : 12.15 ms

        
    

Record the mean for comparison.

Enable shared memory in both configurations

Open both of these working files:

• ~/container_data/ROUTER_CONFIG.json5
• ~/container_data/SESSION_CONFIG.json5

In each file, find the transport/shared_memory block and change its enabled value to:

    

        
        
enabled: true,

    

The router and ROS 2 sessions read these files when their processes start. Stop the router and simulation processes with Ctrl+C, then restart them so they load the updated setting:

    

        
        
# Router terminal
just router

    

    

        
        
# Simulation terminal
just rox_simu use_wall_time:=True no_gui

    

Run the latency measurement again from a sourced terminal:

    

        
        
just cam_latency

    

The supplied reference system produced these results:

The reference mean latency improves by approximately 30%, with lower jitter. Your result depends on the server, workload, and run conditions; compare the two measurements from your own system.

Verify the transport change

Check the shared-memory directory and loopback traffic:

    

        
        
ls /dev/shm
just iftop_lo

    

.zenoh files should appear under /dev/shm for Zenoh processes using shared memory. The large loopback flows should also disappear because the data now moves through memory.

Restore the environment for the next Learning Path

Keep the shared-memory configuration enabled. In Terminal 2, press Ctrl+C to stop the temporary wall-time simulation if it is still running.

Keep the Zenoh router running in Terminal 1. If it has stopped, restart it from a sourced terminal:

    

        
        
source ~/workshop_env.bash
just router

    

In Terminal 2, source the environment if needed and start the normal headless simulation:

    

        
        
source ~/workshop_env.bash
just rox_simu no_gui

    

In Terminal 3, source the environment if needed and restart Navigation2:

    

        
        
source ~/workshop_env.bash
just rox_nav2

    

Confirm that the router is running, the simulation starts without errors, and Navigation2 reports Managed nodes are active. Leave these three processes running for the next Learning Path.

Verify the complete Learning Path

Use this checklist to confirm the final environment:

• Both Docker containers show Up, and their browser desktops are accessible
• The talker and listener continue exchanging messages after the router stops
• just rox_simu no_gui and just rox_nav2 start without errors, and Navigation2 reports Managed nodes are active
• /scan arrives at approximately 8 Hz
• RViz reports Feedback: reached after the robot navigates to a valid goal
• The robot moves through teleoperation or a /cmd_vel publication, and /odom confirms the movement
• just top, just rt_factor, and just iftop_lo produce output
• You recorded both latency measurements, and .zenoh files appear under /dev/shm; the supplied reference system shows lower shared-memory latency

What you’ve accomplished

You’ve built a ROS 2 Jazzy simulation environment on an Arm server, observed Zenoh discovery behaviour, navigated and controlled a Neobotix ROX robot, measured resource use, and compared TCP-loopback communication with shared-memory transport.