Monitoring of OpenWRT

Intro

I have 3 WiFi routers running OpenWrt. My setup is described here. For some time, I’ve been considering adding monitoring for it. Internet searches led me to discover Grafana Cloud with a local Prometheus setup for this purpose. I wanted to experiment with these tools and decided to give it a try. Several excellent articles and resources helped me; thanks to the authors for sharing them! The links I used are below:

• https://grafana.com/blog/2021/02/09/how-i-monitor-my-openwrt-router-with-grafana-cloud-and-prometheus/
• https://gist.github.com/danpawlik/cbec3d9ae1fbf4652a86c6cafc7bd652
• https://www.cloudrocket.at/posts/monitor-openwrt-nodes-with-prometheus/

Below are the main steps and configurations I followed for my setup. I’ve been using it for ~24 hours and love the results. It allows me to understand how the access points are utilized and assess signal quality, enabling necessary adjustments.

Hardware and Software

The setup includes:

• 3 Xiaomi Redmi Router AC2100 devices running OpenWrt v23.05.5 to gather metrics
• A Mac mini with Docker for hosting the Prometheus server and sending metrics to a Grafana Cloud (free account)

Configure OpenWrt

This setup was applied to all 3 routers via SSH.

1. Update the list of packages and install Prometheus exporter packages:

   opkg update
   opkg install prometheus-node-exporter-lua \
   prometheus-node-exporter-lua-nat_traffic \
   prometheus-node-exporter-lua-netstat \
   prometheus-node-exporter-lua-openwrt \
   prometheus-node-exporter-lua-wifi \
   prometheus-node-exporter-lua-wifi_stations
   

2. Edit the /etc/config/prometheus-node-exporter-lua file. The final version looks like this:

   config prometheus-node-exporter-lua 'main'
       option listen_interface 'lan'
       option listen_port '9100'
       option listen_ipv6 '0'
   

   Note: You can check the interface name in the /etc/config/network file.

3. Restart the service to apply the changes:

   /etc/init.d/prometheus-node-exporter-lua restart
   

4. (Optional) Test that you can access the metrics. From a computer on the same network, run:

   curl ROUTER_IP:9100/metrics
   

   The output should look like this:

   

Create Grafana Cloud Account

The next step is to create a Grafana Cloud account (if you don’t already have one) and obtain a Prometheus endpoint with credentials. Once the account is created, go to Connections -> Add new connection -> Hosted Prometheus metrics and generate a new token.

Save the following details for later:

    remote_write:
      - url: https://prometheus-instance.grafana.net/api/prom/push
        basic_auth:
          username: *****
          password: *****

Configure Local Prometheus Server

I have a Mac mini in my network that’s always running, so I’m using it as my Prometheus server for this configuration. It runs inside a Docker container.

1. Create a Prometheus config file (~/tmp/prometheus.yaml). The content of my file looks like this (replace IP_ROUTER_X, Prometheus endpoint, and credentials with the appropriate values):

   global:
     evaluation_interval: "1m"
     scrape_interval: "1m"
     scrape_timeout: "10s"
   scrape_configs:
     - job_name: node
       static_configs:
         - targets:
             - IP_ROUTER_1:9100
             - IP_ROUTER_2:9100
             - IP_ROUTER_3:9100
   remote_write:
     - url: https://prometheus-instance.grafana.net/api/prom/push
       basic_auth:
         username: *****
         password: *****
   

2. Create a Docker volume for metrics:

   docker volume create prometheus-data
   

3. Run the Docker container:

   docker run -d \
       -p 9090:9090 \
       -v ~/tmp/prometheus.yaml:/etc/prometheus/prometheus.yml \
       -v prometheus-data:/prometheus \
       prom/prometheus
   

Configure Dashboard in Grafana Cloud

Within a few minutes, the metrics will appear in Grafana. You can then import or create a new dashboard.

1. Download the OpenWrt Dashboard JSON file from Grafana dashboards or from this GitHub repository.
2. Create a new dashboard:
   • Navigate to Dashboards -> New dashboard -> Import a dashboard.
   • Upload your JSON file.

Your dashboard should now be visible and should look like this:

That’s it! Explore and refine your setup as needed.