Simple and effective tool for measuring ISP performance at home. The tool measures several performance metrics including packet loss, latency, jitter, and DNS performance. It also has an optional speed test to measure bandwidth. Netprobe aggregates these metrics into a common score, which you can use to monitor overall health of your internet connection.
If you'd like to support the development of this project, feel free to buy me a coffee!
https://buymeacoffee.com/plaintextpm
Visit YouTube for a full tutorial on how to install and use Netprobe:
To run Netprobe, you'll need a PC running Docker connected directly to your ISP router. Specifically:
-
Netprobe requires the latest version of Docker. For instructions on installing Docker, see YouTube, it's super easy.
-
Netprobe should be installed on a machine (the 'probe') which has a wired Ethernet connection to your primary ISP router. This ensures the tests are accurately measuring your ISP performance and excluding and interference from your home network. An old PC with Linux installed is a great option for this.
- Clone the repo locally to the probe machine:
git clone https://github.com/plaintextpackets/netprobe_lite.git
- From the cloned folder, use docker compose to launch the app:
docker compose up
- To shut down the app, use docker compose again:
docker compose down
When upgrading between versions, it is best to delete the deployment altogether and restart with the new code. The process is described below.
- Stop Netprobe in Docker and use the -v flag to delete all volumes (warning this deletes old data):
docker compose down -v
- Clone the latest code (or download manually from Github and replace the current files):
git clone https://github.com/plaintextpackets/netprobe_lite.git
- Re-start Netprobe:
docker compose up
-
Navigate to: http://x.x.x.x:3001/d/app/netprobe where x.x.x.x = IP of the probe machine running Docker.
-
Default user / pass is 'admin/admin'. Login to Grafana and set a custom password.
By default the speed test feature is disabled as many users pay for bandwidth usage (e.g. cellular connections). To enable it, edit the .env file to set the option to 'True':
SPEEDTEST_ENABLED="True"
Note: speedtest.net has a limit on how frequently you can connection and run the test. If you set the test to run too frequently, you will receive errors. Recommend leaving the 'SPEEEDTEST_INTERVAL' unchanged.
To change the port that Netprobe Lite is running on, edit the 'compose.yml' file, under the 'grafana' section:
ports:
- '3001:3000'
Change the port on the left to the port you want to access Netprobe Lite on
If the DNS server your network uses is not already monitored, you can add your DNS server IP for testing.
To do so, modify this line in .env:
DNS_NAMESERVER_4_IP="8.8.8.8" # Replace this IP with the DNS server you use at home
Change 8.8.8.8 to the IP of the DNS server you use, then restart the application (docker compose down / docker compose up)
Some users have their own Grafana instance running and would like to ingest Netprobe statistics there rather than running Grafana in Docker. To do this:
- In the compose.yaml file, add a port mapping to the Prometheus deployment config:
prometheus:
...
ports:
- 'XXXX:9090'
... where XXXX is the port you wish to expose Prometheus on your host machine
-
Remove all of the Grafana configuration from the compose.yaml file
-
Run Netprobe and then add a datasource to your existing Grafana as http://x.x.x.x:XXXX where x.x.x.x = IP of the probe machine running Docker
By default, Docker will store the data collected in several Docker volumes, which will persist between restarts.
They are:
netprobe_grafana_data (used to store Grafana user / pw)
netprobe_prometheus_data (used to store time series data)
To clear out old data, you need to stop the app and remove these volumes:
docker compose down
docker volume rm netprobe_grafana_data
docker volume rm netprobe_prometheus_data
When started again the old data should be wiped out.
Using the default method, the data is stored within Docker volumes which you cannot easily access from the host itself. If you'd prefer storing data in mapped folders from the host, follow these instructions (thank you @Jeppedy):
-
Clone the repo
-
Inside the folder create two directories:
mkdir -p data/grafana data/prometheus
- Modify the compose.yml as follows (volume path as well as adding user ID):
prometheus:
restart: always
container_name: netprobe-prometheus
image: "prom/prometheus"
volumes:
- ./config/prometheus/prometheus.yml:/etc/prometheus/prometheus.yml
- ./data/prometheus:/prometheus # modify this to map to the folder you created
command:
- '--config.file=/etc/prometheus/prometheus.yml'
- '--storage.tsdb.path=/prometheus'
networks:
- custom_network # Attach to the custom network
user: "1000" # set this to the desired user with correct permissions to the bind mount
grafana:
restart: always
image: grafana/grafana-enterprise
container_name: netprobe-grafana
volumes:
- ./config/grafana/datasources/automatic.yml:/etc/grafana/provisioning/datasources/automatic.yml
- ./config/grafana/dashboards/main.yml:/etc/grafana/provisioning/dashboards/main.yml
- ./config/grafana/dashboards/netprobe.json:/var/lib/grafana/dashboards/netprobe.json
- ./data/grafana:/var/lib/grafana # modify this to map to the folder you created
ports:
- '3001:3000'
networks:
- custom_network # Attach to the custom network
user: "1000" # set this to the desired user with correct permissions to the bind mount
- Remove the volumes section from compose.yml
Netprobe will automatically restart itself after the host system is rebooted, provided that Docker is also launched on startup. If you want to disable this behavior, modify the 'restart' variables in the compose.yaml file to this:
restart: never
To wipe all stored data and remove the Docker volumes, use this command:
docker compose down -v
This will delete all containers and volumes related to Netprobe.
Q. How do I reset my Grafana password?
A. Delete the docker volume for grafana. This will reset your password but will leave your data:
docker volume rm netprobe_grafana_data
Q. I am running Pihole and when I enter my host IP under 'DNS_NAMESERVER_4_IP=' I receive this error:
The resolution lifetime expired after 5.138 seconds: Server Do53:192.168.0.91@53 answered got a response from ('172.21.0.1', 53) instead of ('192.168.0.91', 53)
A. This is a limitation of Docker. If you are running another DNS server in Docker and want to test it in Netprobe, you need to specify the Docker network gateway IP:
- Stop netprobe but don't wipe it (docker compose down)
- Find the gateway IP of your netprobe-probe container:
$ docker inspect netprobe-probe | grep Gateway
"Gateway": "",
"IPv6Gateway": "",
"Gateway": "192.168.208.1",
"IPv6Gateway": "",
- Enter that IP (e.g. 182.168.208.1) into your .env file for 'DNS_NAMESERVER_4_IP='
Q. I constantly see one of my DNS servers at 5s latency, is this normal?
A. 5s is the timeout for DNS queries in Netprobe Lite. If you see this happening for one specific IP, likely your machine is having issues using that DNS server (and so you shouldn't use it for home use).
This project is released under a custom license that restricts commercial use. You are free to use, modify, and distribute the software for non-commercial purposes. Commercial use of this software is strictly prohibited without prior permission. If you have any questions or wish to use this software commercially, please contact [[email protected]].
