Coreflux MQTT Broker

Overview

This guide walks you through setting up the Coreflux MQTT Broker on a Revolution Pi (ARM64) industrial edge device, deploying LOT (Language of Things) actions from VS Code, and running Python-based vibration severity calculations directly inside the MQTT broker with no cloud dependency.

Data flow

Prerequisites

Hardware

Revolution Pi (ARM64) or any edge device with Ethernet. Learn abou the hardware requirement here: https://docs.coreflux.org/quick-start/installation
Network connection between your dev PC and RevPi
Vibration sensor with IO-Link (or simulated data)

Software (Development PC)

Visual Studio Code
LoT Language Support by Coreflux extension (VS Code marketplace)
Python 3.11.9 from python.org (Not Microsoft Store version, see troubleshooting)

Software (Revolution Pi)

Coreflux MQTT Broker for Linux ARM64
Python 3.11.x (installed via apt)
python3.11-dev package
Node-RED (optional, for visualization)

1. Installing Coreflux on Revolution Pi

Download and Extract

Download the Linux ARM64 build of Coreflux MQTT Broker from coreflux.org. Transfer it to your RevPi via SCP or download directly:

# On RevPi — create install directory
mkdir ~/coreflux
cd ~/coreflux

# Extract the downloaded zip
unzip CorefluxMQTTBroker_linux-arm64_1.10.0.zip

# Make binary executable
chmod +x CorefluxMQTTBroker

Install the Coreflux MQTT Broker

The installation instructions are given on this page: https://docs.coreflux.org/quick-start/installation#raspberry-piStart the broker to verify it works:

cd ~/coreflux
sudo ./CorefluxMQTTBroker

You should see the broker start and listen on port 1883. Press Ctrl+C to stop after confirming startup.

Default CredentialsThe default Coreflux broker credentials are username: root, password: coreflux. These are required when connecting from VS Code or MQTT clients.

Install VS Code

Install the open source code editor i.e., VS Code, on your computer: https://code.visualstudio.com/

Install LoT extension in VS Code

LoT (Language of Things) is a human-readable language for IoT automation. It uses near-English syntax to define logic, data structures, and integrations—all executed directly within the Coreflux MQTT broker.

The LoT extension lets you interact with the Coreflux MQTT broker from VS Code. To install LoT, go to the extension in VS Code and search for LoT.

Learning LoT

To learn how LoT interacts with Coreflux MQTT Broker, check this repository, which has many sample codes and exercises: https://github.com/CorefluxCommunity/Language-Of-Things-Training/tree/main

Configure VS Code Connection

In your LoT project folder, create or edit the .broker file with credentials embedded in the URL:

mqtt://root:coreflux@<RevPi-IP>:1883
#username: root, password: coreflux

Authentication RequiredWithout credentials in the .broker file, the VS Code extension will connect but receive PermissionDenied errors when trying to subscribe to $SYS/Coreflux/ system topics. This causes the ‘Entity not found on broker’ error in the LOT notebook.

2. Setting Up Python Runtime on Revolution Pi

Coreflux uses an embedded Python runtime located in ~/coreflux/python_runtime/. On Linux ARM64, this requires manually copying the system Python shared libraries into this folder.

Install Required Python Packages

sudo apt update
sudo apt install python3.11 python3.11-dev python3.11-full -y

Why python3.11-dev?The -dev package installs the shared library files (libpython3.11.so) that embedded runtimes need to load Python via the C API. Without it, the broker finds the python_runtime folder but cannot initialize Python.

Find the Shared Library Files

find /usr/lib -name "libpython3.11*" 2>/dev/null

# Expected output:
# /usr/lib/aarch64-linux-gnu/libpython3.11.so.1
# /usr/lib/aarch64-linux-gnu/libpython3.11.so.1.0
# /usr/lib/python3.11/config-3.11-aarch64-linux-gnu/libpython3.11.so

Copy Files to python_runtime

# Clear existing contents
sudo rm -rf /home/pi/coreflux/python_runtime/*

# Copy Python executable
sudo cp /usr/bin/python3 /home/pi/coreflux/python_runtime/python

# Copy Python standard library
sudo cp -r /usr/lib/python3.11 /home/pi/coreflux/python_runtime/

# Copy shared libraries (critical — broker loads these at runtime)
sudo cp /usr/lib/aarch64-linux-gnu/libpython3.11.so.1 /home/pi/coreflux/python_runtime/
sudo cp /usr/lib/aarch64-linux-gnu/libpython3.11.so.1.0 /home/pi/coreflux/python_runtime/

Confirm Python Loads at Broker Startup

Start the broker and look for these lines in the output:

sudo ./CorefluxMQTTBroker

# Look for:
# INF  Embedded Python runtime initialized at: /home/pi/coreflux/python_runtime
# INF  Python runtime initialized successfully    <-- this is what we want
# INF  Python Entity Provider initialized

Python WorkingWhen you see ‘Python runtime initialized successfully’ instead of ‘Python runtime not available, scripts disabled’, the Python integration is ready.

3. Troubleshooting

3.1 Python Version Compatibility

Use Python 3.11 OnlyCoreflux 1.10.0 is only compatible with Python 3.11. Using Python 3.12, 3.13, or 3.14 will cause ‘Python runtime not available’ even if all files are correctly copied. Always verify with: python3 —version

3.2 Windows — Microsoft Store Python

On Windows, if Python is installed via the Microsoft Store, the where Python command returns:

C:\Users\YourName\AppData\Local\Microsoft\WindowsApps\python.exe

Microsoft Store Python IncompatibleThe Microsoft Store Python is a stub launcher — it does not contain python311.dll or the Lib/ folder that Coreflux needs. Always install Python from python.org using the Windows installer (64-bit). Download: https://www.python.org/ftp/python/3.11.9/python-3.11.9-amd64.exe

3.3 Windows- Copying Python to python_runtime

After installing Python 3.11.9 from python.org, copy the full installation into the broker folder. This is in case you are running Coreflux Broker on Windows

# First confirm real Python path (should NOT be WindowsApps)
where python

# Copy to broker runtime folder
xcopy /E /I "C:\Users\YourName\AppData\Local\Programs\Python\Python311\*" "C:\Downloads\CorefluxMQTTBroker_win-x64_1.10.0\python_runtime\"

3.4 Port 1883 Already in Use (Linux)

If you see ‘MQTT Port 1883 is in use by unknown process’, another instance of Coreflux is already running (possibly managed by pm2 or systemd):

# Find what is using port 1883
sudo lsof -i :1883

# Kill the process (use the PID shown)
sudo kill -9 <PID>

# If managed by pm2
pm2 stop all

# If managed by systemd
sudo systemctl stop coreflux

3.5 ‘Entity Not Found on Broker’ Error

This error appears in the LoT notebook when clicking the play button, but the action has never been deployed, or the broker has restarted and lost the action. The play button in a .lotnb file deploys the action to the broker; it does not auto-persist between broker restarts unless the action file is saved in the broker’s persistent storage location.

Deployment TipAfter restarting the broker, re-run all cells in your .lotnb file using ‘Run All’ to redeploy all actions.

4. Vibration Severity Example

4.1 The Python Script- VibrationCalc.py

Create this file in your VS Code project. It will be deployed to the broker’s python_scripts folder:

# Script Name: VibrationCalc
import math

# Calculate 3-axis vibration magnitude using Euclidean distance formula
def sqrt_sum_of_squares(x, y, z):
    try:
        result = math.sqrt(float(x)**2 + float(y)**2 + float(z)**2)
        return result
    except (ValueError, TypeError):
        return 0  # Return safe default if sensor sends bad/missing data

ISO 10816 ReferenceThe output of this function is vibration severity in mm/s RMS. According to ISO 10816: Zone A (0-0.7 mm/s) = excellent, Zone B (0.7-1.8) = acceptable, Zone C (1.8-4.5) = monitor, Zone D (>4.5) = danger.

4.2 The LOT Action

Create a new .lotnb file in VS Code with this LOT action code:

DEFINE ACTION OpenPLC
ON EVERY 1 SECONDS DO
    # Pull latest RMS vibration readings from MQTT topics every second
    SET "x_vrms" WITH (GET TOPIC "X_VRMS" AS DOUBLE)
    SET "y_vrms" WITH (GET TOPIC "Y_VRMS" AS DOUBLE)
    SET "z_vrms" WITH (GET TOPIC "Z_VRMS" AS DOUBLE)

    # Hand off all 3 axes to Python and get back a single severity score
    CALL PYTHON "VibrationCalc.sqrt_sum_of_squares"
        WITH ({x_vrms}, {y_vrms}, {z_vrms})
        RETURN AS {vibration_severity}

    # Push the combined severity score out for dashboards/alerts to consume
    PUBLISH TOPIC "output/vibration_severity" WITH {vibration_severity}

4.3 Deploy from VS Code

Open the .lotnb file in VS Code
Confirm the status bar shows ‘MQTT: Connected to mqtt://<RevPi-IP>:1883.’
Enter your user credentials:
1. Username: root
2. Password: coreflux
Click the ▶ play button on the LOT action cell to deploy
Check the COREFLUX ENTITIES panel — the action should appear
Publish test messages to X_VRMS, Y_VRMS, Z_VRMS topics
Subscribe to output/vibration_severity to see results

4.4 Test with MQTT Client

In the following example, we are reading actual vibration data via OPC UA. You can just replace the vibration data source in the flow for your use-case. The source code for this flow is shared in the resource section.

4.5 ISO 10816 Severity Classification Table

Zone	RMS (mm/s)	Classification	Action
Zone	RMS (mm/s)	Classification	Action
A	0 – 0.7	Excellent	New or recently serviced machine
B	0.7 – 1.8	Acceptable	Normal long-term operation
C	1.8 – 4.5	Monitor	Tolerable short-term, investigate
D	> 4.5	Danger	Shutdown recommended immediately

5. Network Configuration on Revolution Pi

5.1 Check Current IPs

hostname -I
ip addr show | grep "inet "

5.1 Assign Static IP to eth1

If you need to assign a static IP to the second Ethernet port (e.g. for direct connection to a PLC or IO-Link master):

# Create network config file
sudo nano /etc/network/interfaces.d/eth1

# Add these lines:
auto eth1
iface eth1 inet static
    address 192.168.100.10
    netmask 255.255.255.0

# If ifup does not apply the config, use ip commands directly:
sudo ip addr flush dev eth1
sudo ip addr add 192.168.100.10/24 dev eth1
sudo ip link set eth1 up

# Verify
ip addr show eth1

6. Managing the Coreflux Broker

6.1 Start with Visible Logs

cd ~/coreflux
sudo ./CorefluxMQTTBroker 2>&1 | tee coreflux.log

6.2 Run in Background (pm2)

# Start with pm2
pm2 start ./CorefluxMQTTBroker --name coreflux
pm2 save

# View status
pm2 status

# View logs
pm2 logs coreflux

# Stop
pm2 stop coreflux

6.3 Run in Background (pm2)

# Check if broker is running and on which port
sudo lsof -i :1883

# Kill specific PID
sudo kill -9 <PID>

# Kill all instances
sudo pkill -f CorefluxMQTTBroker

Resources:

Node-RED flow

coreflux_flow.txt

Quick Reference

Item	Value / Command
Broker port	1883
Default credentials	root / coreflux
.broker file format	mqtt://root:coreflux@<IP>:1883
Python version (required)	3.11.x — do NOT use 3.12+
python_runtime location (Linux)	/home/pi/coreflux/python_runtime/
python_runtime location (Windows)	CorefluxMQTTBroker_win-x64_1.10.0\python_runtime\
Key apt package	python3.11-dev
Check broker port	sudo lsof -i :1883
View broker logs	sudo ./CorefluxMQTTBroker 2>&1 \| tee coreflux.log
Deploy action from VS Code	Click ▶ play button on .lotnb cell
Output topic	output/vibration_severity

Smart platforms

Virtual PLCs

MQTT and edge data

IoT platforms and alerts

Overview

Data flow

Prerequisites

Hardware

Software (Development PC)

Software (Revolution Pi)

1. Installing Coreflux on Revolution Pi

2. Setting Up Python Runtime on Revolution Pi

3. Troubleshooting

3.1 Python Version Compatibility

3.2 Windows — Microsoft Store Python

3.3 Windows- Copying Python to python_runtime

3.4 Port 1883 Already in Use (Linux)

3.5 ‘Entity Not Found on Broker’ Error

4. Vibration Severity Example

4.1 The Python Script- VibrationCalc.py

4.2 The LOT Action

4.3 Deploy from VS Code

4.4 Test with MQTT Client

4.5 ISO 10816 Severity Classification Table

5. Network Configuration on Revolution Pi

5.1 Check Current IPs

5.1 Assign Static IP to eth1

6. Managing the Coreflux Broker

6.1 Start with Visible Logs

6.2 Run in Background (pm2)

6.3 Run in Background (pm2)

Resources:

Node-RED flow

coreflux_flow.txt

Quick Reference

​Overview

​Data flow

​Prerequisites

​Hardware

​Software (Development PC)

​Software (Revolution Pi)

​1. Installing Coreflux on Revolution Pi

​2. Setting Up Python Runtime on Revolution Pi

​3. Troubleshooting

​3.1 Python Version Compatibility

​3.2 Windows — Microsoft Store Python

​3.3 Windows- Copying Python to python_runtime

​3.4 Port 1883 Already in Use (Linux)

​3.5 ‘Entity Not Found on Broker’ Error

​4. Vibration Severity Example

​4.1 The Python Script- VibrationCalc.py

​4.2 The LOT Action

​4.3 Deploy from VS Code

​4.4 Test with MQTT Client

​4.5 ISO 10816 Severity Classification Table

​5. Network Configuration on Revolution Pi

​ 5.1 Check Current IPs

​ 5.1 Assign Static IP to eth1

​6. Managing the Coreflux Broker

​6.1 Start with Visible Logs

​6.2 Run in Background (pm2)

​6.3 Run in Background (pm2)

​Resources:

​Node-RED flow

coreflux_flow.txt

​Quick Reference

Overview

Data flow

Prerequisites

Hardware

Software (Development PC)

Software (Revolution Pi)

1. Installing Coreflux on Revolution Pi

2. Setting Up Python Runtime on Revolution Pi

3. Troubleshooting

3.1 Python Version Compatibility

3.2 Windows — Microsoft Store Python

3.3 Windows- Copying Python to python_runtime

3.4 Port 1883 Already in Use (Linux)

3.5 ‘Entity Not Found on Broker’ Error

4. Vibration Severity Example

4.1 The Python Script- VibrationCalc.py

4.2 The LOT Action

4.3 Deploy from VS Code

4.4 Test with MQTT Client

4.5 ISO 10816 Severity Classification Table

5. Network Configuration on Revolution Pi

5.1 Check Current IPs

5.1 Assign Static IP to eth1

6. Managing the Coreflux Broker

6.1 Start with Visible Logs

6.2 Run in Background (pm2)

6.3 Run in Background (pm2)

Resources:

Node-RED flow

Quick Reference