Monitor temperature with a Raspberry Pi and Grafana/InfluxDB on Docker

This article will detail all the hardware and software I have used to monitor the temperatures from a fridge and freezer to ensure they never go crazy. It comes with notifications as well in case of issues.

Monitor temperature with a Raspberry Pi and Grafana/InfluxDB on Docker

This article will detail all the hardware and software I have used to monitor the temperatures from a fridge and freezer to ensure they never go crazy. It comes with notifications as well in case of issues.

Our target: a neat dashboard in Grafana

First things first, the hardware!

To be able to get this result, you will need:

  1. A Raspberry Pi with Raspbian, it will not consume a lot of power and CPU so Model 1 or 2 are enough
  2. One or more DS18B20 sensors (I'm using 2), depending on how many temperatures you want to capture. They are waterproof and coming with a 2 meters long wire so you can keep your Raspberry Pi far away and safe
  3. A breadboard, it will make your testing and wiring much easier
  4. A 4.7 kOhm resistor
  5. A server to run Docker with Grafana and influxDB, the Raspberry Pi's job is only about capture and push of the data to InfluxDB. You can also run everything on a Raspberry Pi but you will need a more performant one

Hardware installation

Raspberry, resistor and sensors (Source:
  1. As a first step, shut down your Raspberry Pi and plug the sensors and resistor accordingly to the schema above
  2. Open raspi-config with sudo raspi-config
  3. Select Interfacing Options
  4. Highlight 1-wire and set it to Yes then confirm with Select
  5. Reboot your Raspberry Pi with sudo reboot

Sensors configuration

Now we will check if the Raspberry Pi recognize the sensors and can capture temperature

Sensors requirements

  1. You first need to register all the sensors you have plugged to the GPIO4, run sudo modprobe w1-gpio so the Raspberry Pi is aware of it
  2. Now, tell to the Raspberry Pi it can measure temperature on the 1-Wire system with sudo modprobe w1-therm
  3. It's important to load these 2 modules automatically on next boot. For that purpose, edit the modules file with sudo nano /etc/modules and add inside w1-gpioand w1-therm like below
  4. Save with Ctrl+X and Y

Testing the sensors

You need to ensure the sensors are capturing the temperature correctly.

  1. Browse the 1-wire devices folder with cd /sys/bus/w1/devices
  2. List all devices with ls. The DS18B20 sensors start with 28- followed by the serial number of the sensor
  3. Enter in one of the sensors directory and read the file w1_slavewith cd 28-yyyy && cat w1_slave (replace yyyy by your sensor's serial number)
  4. The value you will see at the end of the second line is the current temperature in degrees Celcius (you need to divide it by 1000, t=-16812 means -16.812°C)
  5. The first line contain a CRC, if the value is not YES, you might have a defect in your sensor or in its wiring
Checking the sensor value

Prepare InfluxDB and Grafana to receive the Data

Build the stack

Let's build Grafana and InfluxDB as a docker-compose stack.

version: "2.3"
    image: influxdb:1.8-alpine
    container_name: monitoring-temperature-influxdb
    restart: always
      - "8083:8083"
      - "8086:8086"
      - "8090:8090"
      - PUID=1026
      - PGID=1000
      - TZ=Europe/Paris
      - /home/thelazyfox/docker/monitoring-temperature/influxdb/env.influxdb
      - /home/thelazyfox/docker/monitoring-temperature/influxdb/data:/var/lib/influxdb

    image: grafana/grafana
    container_name: monitoring-temperature-grafana
    restart: always
      - "3000:3000"
      - influxdb
      - /home/thelazyfox/docker/monitoring-temperature/grafana/data:/var/lib/grafana
      - PUID=1026
      - PGID=1000
      - TZ=Europe/Paris
      - GF_LOG_FILTERS=rendering:debug
      - GF_INSTALL_PLUGINS=grafana-clock-panel,briangann-gauge-panel,natel-plotly-panel,grafana-simple-json-datasource

    image: grafana/grafana-image-renderer
    container_name: monitoring-temperature-grafana-image-renderer
    restart: always
      - "8123:8081"
      - PUID=1026
      - PGID=1000
      - TZ=Europe/Paris
      - LOG_LEVEL=debug
      - RENDERING_MODE=default
      - ENABLE_METRICS=true

Please remember to create all folders and files before trying to launch your docker-compose.

How does it look like on my server:

      |_ docker-compose.yaml
      |_ influxdb/
          |_ env.influxdb/ 
          |_ data/ 
      |_ grafana/          
          |_ data/ 

Few notes regarding the stack:

  • grafana-image-renderer is used to integrate charts screenshot in the notifications
  • grafana-image-renderer IP and port must be updated into GF_RENDERING_SERVER_URL environment variable of grafana service
  • Grafana's IP and port must be updated into GF_RENDERING_CALLBACK_URL environment variable of grafana service

And now, launch your docker-compose file with the following command cd monitoring-temperature && docker-compose up -d

Configure InfluxDB

It's time to create your InfluxDB database.

  1. Enter in your container with docker exec -it monitoring-temperature-influxdb bash
  2. Open InfluxDB shell with influx
  3. Create your database named "temp_logger_db" with CREATE DATABASE temp_logger_db
  4. Confirm it was successfully created by running SHOW DATABASES
  5. You can now enter exit and exit again to resume
Database creation in InfluxDB via CLI

Configure Grafana

Grafana is accessible on port 3000, login with the default account admin:admin to link InfluxDB as a data source into Grafana.

  1. Open Settings
  2. Select "Data Sources" tab
  3. Click on "Add data source"
  4. Search and select "InfluxDB"
  5. Configure as below (Server URL, Database name and credentials)
InfluxDB configuration in Grafana

Capture data with Python3 on Raspberry Pi

InfluxDB is ready and Grafana is connected to InfluxDB. Now it's time to go back on your Raspberry Pi to capture sensors' data from the files we saw before and inject them into InfluxDB.

Install dependencies

You need to install python3 and its influx library with sudo apt-get install python3 python3-pip && pip install influxdb

Create the script

  1. To capture the data, you have to create the python script below with sudo nano
  2. Replace the sensors serial numbers on line 17 and integrate your InfluxDB information on line 20 to 23 and save
  3. Test the script by running /usr/bin/python3 /home/pi/ -db=temp_logger_db -sn=data
  4. You should see your temperatures
# -*- coding: utf-8 -*-
import os
import glob
import argparse
import time
import datetime
import sys
from influxdb import InfluxDBClient

os.system('modprobe w1-gpio')
os.system('modprobe w1-therm')

base_dir = '/sys/bus/w1/devices/'
device_folders = glob.glob(base_dir + '28*')

# List of temperature sensors from base_dir. Add as much as you want.
sensors=['28-012032d0174e', '28-012032e0a6f0']

# InfluxDB parameters
host = ""
port = 8086
user = "root"
password = "root"

def get_args():
    parser = argparse.ArgumentParser(description='Push Celsius temperatures from DS18B20 sensors to an Influx database.')
        '-db', '--database', type=str, help='Database name', required=True)
        '-sn', '--session', type=str, help='Session', required=True)
    now =
        '-rn', '--run', type=str, help='Run number', required=False,
    args = parser.parse_args()
    db_name = args.database
    run_no =
    session = args.session
    return db_name, session, run_no

def read_temperature_raw(device_file):
    f = open(device_file, 'r')
    lines = f.readlines()
    return lines

def read_temperature(device_file):
    lines = read_temperature_raw(device_file)

    # Wait for the file to be available with a valid CRC
    while (not lines) or (lines[0].strip()[-3:] != 'YES'):
        print ("Retrying...")
        lines = read_temperature_raw(device_file)

    # Get the temperature value from the file
    equals_pos = lines[1].find('t=')

    # Convert to Celcius and round to 1 decimal
    if equals_pos != -1:
        temperature_string = lines[1][equals_pos+2:]
        temperature = float(temperature_string) / 1000.0
        temperature = round(temperature, 1)

    return temperature

def get_data_points():
    # Go through all sensors to get the temperature
    for sensor in range (len(sensors)):
        device_file=device_folders[sensor]+ '/w1_slave'
        sensors[sensor] = read_temperature(device_file)
        print (device_file,sensor,sensors[sensor])

    # Create Influxdb datapoints
    datapoints = [
            "measurement": session,
            "tags": {"runNum": run_no,},
            "time": timestamp,
            "fields": {"temperature 1":sensors[0],"temperature 2":sensors[1]}
    return datapoints

# Get values from arguments
db_name, session, run_no = get_args()
print ("Session: ", session)
print ("Run No: ", run_no)
print ("DB name: ", db_name)

# Initialize the Influxdb client
client = InfluxDBClient(host, port, user, password, db_name)

# Collecting and pushing data
print("Write points {0} Bresult:{1}".format(datapoints, bResult))

Run it automatically as a cron job

You want to capture the data as soon as the Raspberry Pi is turned on and on a regular basis, for that purpose, we need to create a cronjob.

  1. Open the crontab with crontab -e
  2. Add this line to run the script every minute: * * * * * /usr/bin/python3 /home/pi/ -db=temp_logger_db -sn=data
    If you need help with crontab to change the frequency, you can use cronhub
  3. Update the script's path and the database name if needed
  4. Save with Ctrl+X and Y
  5. It's over!
Crontab file

Run it as a service with systemd

As an alternative, you can also manage your data capture with a systemd service. First you will need to slightly alter the script.

Adapt the script

Replace the section "Collecting and pushing data" from the previous script with this one:

# Collecting and pushing data
     while True:
        print("Write points {0} Bresult:{1}".format(datapoints,bResult))
except KeyboardInterrupt:
    print ("Program stopped by keyboard interrupt [CTRL_C].")
Replace the value "60" by any value in seconds, it will define the data capture cadence

Create and start the service

  1. Create the file with nano /etc/systemd/system/templogger.service
  2. Copy/Paste the service below and update the script's path and the database name if needed
  3. Save with Ctrl+X and Y
  4. Reload the list of systemd scripts with sudo systemctl daemon-reload
  5. Start the service with sudo systemctl start templogger.service
  6. Enable it so it runs automatically on boot with sudo systemctl enable templogger.service
Description=Temp Logger

ExecStart=/usr/bin/python3 /home/pi/ -db=temp_logger_db -sn=data



✌🏼 By crontab or by service, your temperatures are now captured every 60 seconds in InfluxDB! You can now start building your Grafana Dashboard.

Start building your Grafana Dashboard

  1. Create a new Dashboard
  2. Add a new Panel and configure the Query as below
  3. Source InfluxDB from default, the dataset is data (you can customize it by changing the value of -sn in the command)
  4. Select field(temperature1) (you will see one value per sensor)
  5. Format as Time series
  6. Save
Panel configuration in Grafana

🎉 Congrats! Your Raspberry Pi is now capturing temperatures and pushing it in an InfluxDB available through Grafana!
Feel free to share your Dashboards and alerts on Discord!

Since I initially wrote this article, I have printed my own board to replace the breadboard. I've designed it on EasyEDA. I put it for download below, feel free to use it and print it with

Looking for help?
If you are looking for some help or want a quick chat, please head over to the Discord Community!