Published January 15, 2021 © GPL3+

MONITORING RASPBERRY PIs

A simple interface to monitor my raspberry pis (headless) and my laptop thanks to MQTT and Node Red

BeginnerShowcase (no instructions)2 hours2,360

Things used in this project

Hardware components

Raspberry Pi 4 Model B

Software apps and online services

Node-RED

MQTT

Story

Happy owner of a pihole based on a Raspberry Pi zero W and a Raspberry Pi 4 which is used as a server for several purposes (owncloud, MQTT broker, back up of my laptop...) I wanted to be able to easily monitor them with a webbased interface.

So I chose to develop the following solution :

Each board or laptop is running a python script which gather information such as cpu speed, core temperature, disk usage... and compact them into a json message send them to my MQTT Broker
The Raspberry Pi 4 (which is also the MQTT Broker) uses this information through a node red flow to build a nice (in my opinion) dashboard.

The overall project is quite simple and was especially a good way to discover json format, mqtt and basic node red usage.

"Client" Side

On the client side the code running is based on Python 3.

It gathers the following values thanks to psutils lib :

Name of the host
Uptime
Core Temperature
Virtual Memory (%)
Disque usage (%)
Average load (1min, 5min, 15min)
CPU Usage per core (%)

The value are next packed into a json message and published to MQTT broker.

To enable the.py script to run at boot, I had it to my cron tab (@reboot...).

"Server" Side

As previously said the server is running thanks to Node Red. It does the following actions :

Receiving MQTT message
Print if client is connected or not
Decoding Json message
Printing on the dash board the different value (graph, bar plot,...)
Sending an email if the disk usage is over 90%

I used the dashboard nodes and a node called "artless gauges"

To access to dashboard, I just connect to <host_IP>:1880/ui

Example of a subflow for one of the client :

Subflow detail

Dashboard :

Focus on some nodes and functions :

JSON received :

example of Json message

Example of JSON decoding before printing

Example of the disk usage node to send an email if du > 90%:

Translation :

msg.topic="Alert disk space Odysseus"

msg.payload="Attention : Odysseus's disk usage is over 90%"

Python code - Client Side

#! /usr/bin/env python3

import psutil
import time
import datetime
import json
import paho.mqtt.client as mqttClient
name = ""

broker = ""
port = 1883
user = ""
password = ""


def on_connect(client, userdata, flags, rc):
    if rc == 0:
        print("Connected to broker")
        global Connected  # Use global variable
        Connected = True  # Signal connection
    else:
        print("Connection failed")


client = mqttClient.Client()  # create new instance
client.username_pw_set(user, password=password)  # set username and password
client.on_connect = on_connect  # attach function to callback
client.connect(broker, port=port)  # connect to broker

client.loop_start()  # start the loop
i = 0
while 1:
    temperature_main = psutil.sensors_temperatures()["asus"][0][1]
    du = psutil.disk_usage("/")[3]
    virtual_memory = psutil.virtual_memory().percent
    cpu_percent = psutil.cpu_percent(percpu=True)
    cpu_load_avg = psutil.getloadavg()
    uptime = str(datetime.datetime.now().replace(microsecond=0) - datetime.datetime.fromtimestamp(
        psutil.boot_time()).replace(microsecond=0))
    messageToSend = {"name": name, "du": du, "virtual_memory": virtual_memory, "temperature_main": temperature_main, "cpu_percent": cpu_percent, "cpu_load_avg": cpu_load_avg, "uptime": uptime}
    messageToSend_JSON = json.dumps(messageToSend)
    client.publish("topic", messageToSend_JSON)
    print("Message #", i, "published to", broker)
    i+=1
    time.sleep(1)

Credits

aerodynamics

4 projects • 27 followers

MONITORING RASPBERRY PIs

Things used in this project

Hardware components

Software apps and online services

Story

"Client" Side

"Server" Side

Focus on some nodes and functions :

Code

Python code - Client Side

Credits

aerodynamics

Comments

Embed the widget on your own site

MONITORING RASPBERRY PIs

MONITORING RASPBERRY PIs

Things used in this project

Hardware components

Software apps and online services

Story

"Client" Side

"Server" Side

Focus on some nodes and functions :

Code

Python code - Client Side

Credits

aerodynamics

Comments

Related channels and tags