APT monitoring with Prometheus

This repository contains all the assets to monitor APT upgrades and security upgrades from a Linux system using Prometheus. It also tells you if a system needs a reboot.

Requirements

• Node Exporter is deployed. I have tested it with Node_Exporter 1.5.0.
• Text file collector is enabled.
• Packages moreutils and update-notifier-common for sponge and the apt-check command.

Usage

Step 1: Install sponge and apt-check dependency packages

sudo apt install moreutils update-notifier-common

Step 2: Deploy metric wrapper script to /usr/share/apt-metrics.sh

sudo wget https://raw.githubusercontent.com/labmonkeys-space/apt-prometheus/main/script/apt-metrics.sh -O /usr/share/apt-metrics.sh
sudo chmod +x /usr/share/apt-metrics.sh

Step 3: Install APT hook to /etc/apt/apt.conf.d

sudo wget https://raw.githubusercontent.com/labmonkeys-space/apt-prometheus/main/apt.conf.d/60prometheus -O /etc/apt/apt.conf.d/60prometheus

Step 4: Configure Node_Exporter to fetch content from a textfile directory

The APT hook writes the apt.prom file to /var/lib/node_exporter/textfile_collector.

Configure your Node_Exporter to pick up the text file from the directory. The text file collector is enabled with --collector.textfile and the lookup directory is configured in --collector.textfile.directory. Here is my example from the Node_Exporter ExecStart arguments defined in the Systemd unit.

--collector.textfile --collector.textfile.directory=/var/lib/node_exporter/textfile_collector

Credits

Credits to Tom Henderson who did the heavy lifting in the apt-metrics.sh script. He also gives some hints how to annotate the Grafana dashboard. I've skipped this part and was not so helpful in my environment.