Building a Self-Hosted, Privacy-First Static Site Analytics Platform on a VPS with Umami and ClickHouse

Introduction: The Shift Toward Data Sovereignty

In the modern digital ecosystem, tracking user engagement while respecting data privacy has become a paramount challenge for businesses and developers alike. Standard solutions like Google Analytics 4 (GA4) have increasingly come under scrutiny due to complex GDPR/CCPA compliance hurdles, bloated client-side scripts, and data sovereignty concerns. For businesses operating static websites, blogs, or high-traffic landing pages, the need for a lightweight, privacy-first, and self-hosted analytics infrastructure has never been more urgent.

This comprehensive guide outlines how to architect and deploy an enterprise-grade tracking platform on a single Virtual Private Server (VPS). By combining Umami, an open-source, privacy-focused analytics frontend, with ClickHouse, a hyper-fast column-oriented database management system, you can capture millions of events with minimal resource overhead and total data ownership.

Why Umami and ClickHouse?

While Umami natively supports traditional relational databases like PostgreSQL or MySQL, scaling to millions of monthly pageviews on a budget-friendly VPS requires a specialized storage engine. This is where ClickHouse excels.

Umami Analytics: A clean, modern alternative to Google Analytics. It does not use cookies, does not collect personally identifiable information (PII), anonymizes visitor IPs automatically, and loads a tiny 2KB script that won't degrade your Core Web Vitals.
ClickHouse DB: A columnar database designed specifically for Online Analytical Processing (OLAP). Instead of processing rows, ClickHouse reads columns, allowing it to aggregate billions of rows in milliseconds while utilizing advanced data compression algorithms to save up to 90% on disk space.

Combining these two technologies allows a modest 2 vCPU / 4GB RAM VPS to easily handle traffic loads that would traditionally require expensive, enterprise-tier SaaS subscriptions.

Prerequisites and Architecture Overview

Before initiating the deployment, ensure you have access to a clean Ubuntu VPS (minimum recommended specs: 2 vCPUs, 4GB RAM, and 40GB SSD) with Docker and Docker Compose installed. You will also need a domain name pointed to your VPS IP address for SSL termination via a reverse proxy.

Our architectural blueprint relies on a containerized environment managed via Docker Compose:

Nginx Proxy Manager / Caddy: Handles incoming HTTPS requests and manages SSL certificates.
Umami Web Service: The Node.js application hosting the dashboard and ingestion API.
ClickHouse Server: The analytical powerhouse storing tracking events.

Step-by-Step Deployment Guide

Step 1: Preparing the ClickHouse Environment

First, we need to create the directory structure and initialize configuration files for ClickHouse to ensure it is secure and optimized for analytical workloads. Create a dedicated project directory and navigate into it:

mkdir -p self-hosted-analytics/clickhouse && cd self-hosted-analytics

Within the clickhouse folder, create a customized users.xml file to set secure credentials. ClickHouse relies on tight user access controls to prevent unauthorized query execution.

Step 2: Configuring Docker Compose

We will utilize Docker Compose to orchestrate our stack. Create a docker-compose.yml file in your root project folder. This file defines the network isolation and environment variables required for Umami to communicate seamlessly with ClickHouse.

Crucially, you must specify the DATABASE_URL environment variable using the specific ClickHouse connection string format:

CLICKHOUSE_URL=clickhouse://username:password@clickhouse-server:8123/umami

This directs Umami's Prisma ORM to route ingestion and analytical queries away from the standard relational engine and directly into ClickHouse's high-performance tables.

Step 3: Launching the Stack and Setting Up the Reverse Proxy

With your configurations defined, execute the deployment command:

docker compose up -d

Verify that all containers are healthy by checking the logs. Once confirmed, configure your reverse proxy (such as Nginx or Caddy) to route external traffic from your sub-domain (e.g., analytics.yourcompany.com) to internal port 3000, ensuring SSL encryption is strictly enforced.

Integrating the Privacy-First Script into Your Static Site

Once logged into the Umami dashboard, create a new website profile. Umami will generate a lightweight tracking script. Because it is privacy-first, you do not need to display annoying cookie consent banners to your users; the script operates entirely without cookies or persistent local storage identifiers.

Insert the following snippet into the tag of your static site configuration (e.g., Astro, Next.js, Hugo, or Gatsby):

Optimization Tip: To prevent ad-blockers from aggressively blocking your analytics script, you can easily use Umami’s environment variables to proxy or rename the script.js file to a custom name like metrics.js.

Performance Tuning and Long-Term Maintenance

Running analytics on a VPS requires strategic optimization to maintain stability during traffic spikes:

Memory Limits: Bound ClickHouse’s maximum memory consumption via its configuration files to prevent the Linux Out-Of-Memory (OOM) killer from terminating the process.
Data Retention Policies: Since ClickHouse compresses data exceptionally well, you can retain raw event data for years. However, setting up a 12-month TTL (Time-To-Live) policy ensures your disk space remains highly predictable.
Automated Backups: Schedule automated nightly backups of your ClickHouse metadata and data directories using tools like clickhouse-backup to an off-site S3-compatible cloud storage bucket.

Conclusion: Complete Data Ownership

By migrating from third-party tracking services to a self-hosted Umami and ClickHouse stack, your business gains a massive competitive advantage. You drastically reduce client-side loading latency, completely bypass the regulatory compliance complexities associated with third-party data sharing, and maintain 100% data sovereignty over your user engagement metrics. All of this is achieved on an economical VPS infrastructure designed to scale smoothly alongside your enterprise.