Pulse — Real-Time Monitoring Dashboard for Proxmox, Docker, and Kubernetes

Introduction

Pulse is a self-hosted monitoring dashboard that brings Proxmox VEs, Docker containers, and Kubernetes clusters into a single, real-time view. It targets homelab operators and small infrastructure teams who want unified visibility without deploying a full observability stack like Prometheus and Grafana. Pulse connects directly to your infrastructure APIs and presents live metrics in a clean, modern dashboard.

What Pulse Does

• Displays real-time CPU, memory, disk, and network metrics for Proxmox VMs, LXC containers, and hosts
• Monitors Docker container status, resource usage, and logs across multiple Docker hosts
• Provides Kubernetes cluster visibility with pod, node, and workload health summaries
• Sends smart alerts via webhooks when resources cross configurable thresholds
• Offers AI-powered diagnostic suggestions to help identify root causes of performance issues

Architecture Overview

Pulse is built with Go on the back-end and React with TypeScript on the front-end. The server polls infrastructure APIs (Proxmox REST API, Docker Engine API, Kubernetes API) at configurable intervals and pushes updates to connected browsers via WebSocket. Metric history is stored in an embedded time-series store for short-term trend visualization. The single binary serves both the API and the static front-end, keeping deployment simple.

Self-Hosting & Configuration

• Run the Docker image with environment variables pointing to your Proxmox, Docker, or Kubernetes endpoints
• Create a Proxmox API token with PVEAuditor role for read-only monitoring access
• For Docker monitoring, mount the Docker socket or configure TCP access with TLS
• Set alert thresholds and webhook URLs for Slack, Discord, or generic HTTP notifications
• Deploy behind a reverse proxy with TLS for secure remote dashboard access

Key Features

• Unified dashboard covering Proxmox, Docker, and Kubernetes from a single pane of glass
• Real-time metric streaming via WebSocket with no page refresh required
• Smart alerting with configurable thresholds for CPU, memory, disk, and container health
• AI-powered insights that suggest potential causes when anomalies are detected
• Minimal resource footprint suitable for running alongside existing homelab workloads

Comparison with Similar Tools

• Grafana + Prometheus — powerful and extensible but requires deploying and configuring multiple components; Pulse is a single container with zero setup
• Netdata — comprehensive per-host monitoring agent; Pulse aggregates across multiple hosts and platforms in one view
• Beszel — lightweight server monitoring; Pulse adds Proxmox-native and Kubernetes integration
• Uptime Kuma — focuses on endpoint availability checks; Pulse monitors infrastructure resource metrics rather than HTTP uptime
• Proxmox built-in — basic graphs per node; Pulse provides a cross-node, cross-platform unified view with alerting

FAQ

Q: Do I need Prometheus or any external time-series database? A: No. Pulse includes an embedded metric store for short-term history. It connects directly to Proxmox, Docker, and Kubernetes APIs without intermediaries.

Q: Can I monitor multiple Proxmox clusters? A: Yes. You can configure multiple Proxmox endpoints, and Pulse aggregates them into a single dashboard view.

Q: What permissions does the Proxmox API token need? A: A token with the PVEAuditor role provides sufficient read-only access for monitoring VMs, containers, and host metrics.

Q: Is there a mobile app? A: Pulse is a responsive web application. There is no native mobile app, but the dashboard works well in mobile browsers.

Sources

• https://github.com/rcourtman/Pulse
• https://pulseapp.dev