GoAccess vs Umami: Self-Hosted Analytics Compared

Quick Verdict

Want analytics without touching your site’s code? GoAccess reads your existing server logs — no tracking scripts, no cookies, zero page load impact. Want a polished Google Analytics replacement with real-time dashboards and event tracking? Umami gives you that with a lightweight JavaScript snippet. For most self-hosters running Nginx or Apache, GoAccess is the simpler path. For anyone building a product or tracking user behavior across pages, Umami is the better tool.

Updated March 2026: Verified with latest Docker images and configurations.

Overview

GoAccess is a real-time web log analyzer written in C. It parses Nginx, Apache, Caddy, or any CLF/combined log format and generates analytics in your terminal or as a self-updating HTML report. No JavaScript. No cookies. No database. GoAccess reads what your web server already records. Official site.

Umami is a JavaScript-based analytics platform built with Next.js. You add a <script> tag to your pages, and it collects page views, referrers, browser data, screen sizes, and custom events — all stored in PostgreSQL or MySQL. Cookie-free by design and GDPR-compliant without consent banners. Official site.

These tools solve the same problem (web analytics) in fundamentally different ways. GoAccess is passive — it analyzes data your server already generates. Umami is active — it requires instrumentation but captures richer data.

Feature Comparison

Feature	GoAccess	Umami
Data source	Server access logs (Nginx, Apache, Caddy)	JavaScript tracking snippet
Cookies	None (log-based)	None (cookieless by design)
Page load impact	Zero	Minimal (~2 KB script)
Real-time dashboard	Terminal UI + WebSocket HTML report	Web dashboard with live updates
Custom event tracking	No	Yes (JavaScript API)
Multi-site support	One log source per instance	Unlimited websites per instance
User sessions	Estimated from IP + UA	Tracked with anonymous session IDs
API	None (HTML report export)	Full REST API
Database	Flat file (optional persistent storage)	PostgreSQL or MySQL required
Geolocation	GeoIP database (MaxMind)	Built-in country detection
Bot filtering	Configurable exclusion patterns	Automatic bot detection
Data retention	Configurable (or keep all logs)	Configurable per-site
License	MIT	MIT

Installation Complexity

GoAccess needs one container and access to your web server’s log files:

services:
  goaccess:
    image: allinurl/goaccess:1.10.1
    container_name: goaccess
    restart: unless-stopped
    ports:
      - "7890:7890"
    volumes:
      - /var/log/nginx/access.log:/var/log/access.log:ro
      - goaccess-data:/srv/data
      - goaccess-html:/srv/report
    command: >
      --log-file=/var/log/access.log
      --log-format=COMBINED
      --real-time-html
      --ws-url=wss://analytics.example.com
      --output=/srv/report/index.html
      --persist
      --restore
      --db-path=/srv/data

volumes:
  goaccess-data:
  goaccess-html:

No database container. No config files. Point it at your log file and it works.

Umami requires two containers — the app and a PostgreSQL database:

services:
  umami:
    image: ghcr.io/umami-software/umami:v3.0.3
    container_name: umami
    restart: unless-stopped
    ports:
      - "3000:3000"
    environment:
      DATABASE_URL: "postgresql://umami:changeme@umami-db:5432/umami"
      APP_SECRET: "change-me-to-random-string"
    depends_on:
      umami-db:
        condition: service_healthy

  umami-db:
    image: postgres:17
    container_name: umami-db
    restart: unless-stopped
    environment:
      POSTGRES_DB: umami
      POSTGRES_USER: umami
      POSTGRES_PASSWORD: changeme
    volumes:
      - umami_db:/var/lib/postgresql/data
    healthcheck:
      test: ["CMD-SHELL", "pg_isready -U umami"]
      interval: 10s
      timeout: 5s
      retries: 5

volumes:
  umami_db:

After deploying Umami, you still need to add the tracking script to every page on your site. GoAccess requires zero changes to your website.

Winner: GoAccess. One container, no database, no code changes. Plug into your existing logs and go.

Performance and Resource Usage

Metric	GoAccess	Umami
RAM (idle)	30–80 MB	150–300 MB (app + PostgreSQL)
Disk (30 days, 10K pages/day)	5–20 MB (depends on log retention)	50–200 MB (PostgreSQL storage)
CPU	Spikes when parsing large logs	Low (async event processing)
Network overhead	None	~2 KB per page view (tracking script)
Containers	1	2 (app + database)

GoAccess is dramatically lighter. It processes logs with zero network overhead and can run alongside your web server with minimal resource impact. Umami’s PostgreSQL database is the main resource consumer — acceptable for most servers, but unnecessary if you only need basic visitor stats.

Data Accuracy

GoAccess counts every HTTP request your server handles, including bots, crawlers, and prefetch requests. You configure exclusion patterns manually. This means raw numbers from GoAccess tend to be higher than real visitor counts.

Umami only counts visits where JavaScript executes in a real browser. Bots that don’t run JavaScript are invisible. Ad blockers that block the tracking script will undercount real visitors. Neither tool uses cookies, so both estimate unique visitors rather than tracking them precisely.

Accuracy dimension	GoAccess	Umami
Bot inflation	Higher (needs manual filtering)	Lower (JS-only)
Ad blocker impact	None (server-side)	Undercounts blocked visitors
JavaScript-disabled visitors	Counted	Missed
Single-page app tracking	No (only initial page load logged)	Yes (with router integration)

For content sites with server-rendered pages, GoAccess is more complete. For SPAs and interactive products, Umami captures user behavior that never hits server logs.

Use Cases

Choose GoAccess If…

• You want analytics without adding tracking scripts to your site
• Privacy is paramount — you want zero client-side data collection
• You already run Nginx, Apache, or Caddy and want to use existing logs
• You prefer a terminal-first workflow
• Resource efficiency matters — you can’t spare RAM for PostgreSQL
• You run a static site and just need traffic numbers

Choose Umami If…

• You need custom event tracking (button clicks, form submissions, signups)
• You manage multiple websites from one dashboard
• You want a clean web UI similar to Google Analytics
• Your team needs an API to pull analytics data programmatically
• You run a JavaScript-heavy SPA where server logs miss client-side navigation
• You want shareable public dashboards for stakeholders

Community and Support

Dimension	GoAccess	Umami
GitHub stars	~19,000	~24,000
First release	2009	2020
Primary language	C	TypeScript/Next.js
Development pace	Stable, infrequent releases	Active, frequent releases
Documentation	Good (man page + website)	Excellent (docs.umami.is)
Hosted option	No	Yes (Umami Cloud)

GoAccess has been around since 2009 and is rock-solid — it rarely needs updates because log parsing doesn’t change. Umami moves faster with new features (custom reports, funnel analysis, UTM tracking) but may require more frequent upgrades.

Final Verdict

The practical choice is GoAccess because it requires zero changes to your website and zero additional infrastructure. If you already have a web server generating logs, you’re five minutes from analytics. No tracking scripts to maintain, no database to back up, no JavaScript for visitors to block.

Choose Umami when you outgrow log-based analytics — when you need event tracking, multi-site management, team access, or API integrations. It is a genuine Google Analytics replacement, not just a traffic counter.

For many self-hosters, running both makes sense: GoAccess for server-level traffic monitoring, Umami for user-facing analytics you share with others.

FAQ

Can GoAccess track clicks or form submissions?

No. GoAccess only analyzes HTTP requests logged by your web server. Client-side interactions that don’t generate a new HTTP request are invisible. Use Umami or Plausible for event tracking.

Does Umami work with static sites?

Yes. Add the tracking script tag to your HTML template. It works with Hugo, Astro, Jekyll, or any static site generator. The script is under 2 KB and doesn’t slow page loads.

Which is more GDPR-compliant?

Both are cookie-free and can be GDPR-compliant. GoAccess has an edge because it collects no client-side data at all — everything comes from server logs that your web server already generates. Umami’s JavaScript snippet technically collects browser data (screen size, language, referrer), though it anonymizes it.

Can I use GoAccess with Cloudflare or CDNs?

Partially. If Cloudflare serves cached pages, those requests never hit your origin server and won’t appear in your logs. You can use Cloudflare’s CF-Connecting-IP header in your log format, but cached-page visits will be undercounted.

Related

• How to Self-Host GoAccess
• How to Self-Host Umami
• Plausible vs Umami
• GoAccess vs Plausible
• Umami vs Matomo
• Best Self-Hosted Analytics
• Self-Hosted Alternatives to Google Analytics