kube-inventory 1.0.1

Inventory of softwares running in a Kubernetes cluster

Kube Inventory

A tool to inventory and track software versions running in Kubernetes clusters, helping identify outdated applications that may need upgrades.

Overview

Kube Inventory scans your Kubernetes cluster to identify running applications and compares their versions against the latest releases available in their GitHub repositories. This helps you quickly assess which applications need updates.

Features

• Multiple Output Formats: Export inventory data as json, csv or expose via Prometheus metrics
• GitHub Integration: Automatically fetches latest release information from GitHub repositories
• Kubernetes Native: Works with standard kubeconfig or in-cluster configuration
• Custom Application Support: Extend the default application list with custom definitions
• Prometheus Integration: Expose metrics for monitoring and alerting
• Web Interface: Built-in web server for Prometheus mode

Utilization

The tool can be run from the command line with various options to customize its behavior or by Kubernetes, and then generating Prometheus metrics for monitoring purposes.

CLI Mode

You can run Kube Inventory from the command line to generate inventory reports in your desired format. Use the --output-mode flag to specify the format (json or csv).

The installation can be easily done via pip:

pip install kube-inventory

Example command to run Kube Inventory in CLI mode:

kube-inventory

The reports will be saved in the default output directory specified by the argument --output-dir, which defaults to the current directory.

In this sample output, you can see the generation of the report in the csv format:

Prometheus Mode

You can also run Kube Inventory in Prometheus mode, where it exposes the inventory data as metrics via a built-in web server. This allows you to integrate with Prometheus for monitoring and alerting.

You can specify the output mode as prometheus using the --output-mode flag, and them application will start a web server to expose the metrics in Prometheus format.

In case you want to install Kube Inventory in a Kubernetes cluster, you can use the provided Helm chart available in the srekit/helm-charts/kube-inventory,

You could also have the possibility to visualize the metrics using Grafana dashboards, besides trigger alerts using Prometheus rules.

Command-Line Options

Argument	Options	Description	Default
--extra-apps-file-path	Path to YAML file	Path to custom applications YAML file	None
--github-access-token	String	GitHub access token for API requests	None
--github-api-url	URL	Base URL for GitHub API	https://api.github.com
--kube-config-path	Path to kubeconfig file	Path to the kubeconfig file to use	~/.kube/config
--kube-in-cluster	Boolean	Use in-cluster Kubernetes config	False
--log-level	debug, info, warning, error	Logging level for the application	info
--output-dir	Path to output directory	Directory to save output files	Current directory
--output-mode	json, csv, prometheus	Output format for the inventory data	json
--output-refresh-interval-seconds	Integer (seconds)	Refresh interval for Prometheus mode	300
--version		Show application version and exit
--web-host	Hostname or IP	Host for the web server in Prometheus mode	0.0.0.0
--web-port	Integer (port number)	Port for the web server in Prometheus mode	8080

Environment Variables

The following environment variables can be used as alternatives to command-line arguments:

Environment Variable	Description
GITHUB_ACCESS_TOKEN	GitHub access token for API requests
LOG_LEVEL	Logging level for the application