Hosting Activepieces: on your own VPS or managed, hosted in Germany (EU)

What Activepieces is (and what "Community Edition" means)

Activepieces is a visual workflow-automation tool: triggers (webhook, schedule, app event) start flows made of steps from ready-made integrations ("pieces", listed on activepieces.com/pieces), HTTP requests or your own code. According to the Activepieces Docs: License page and the repository LICENSE (GitHub) (both read 5 November 2026), the core is released under the MIT licence; only the enterprise code in packages/ee and packages/server/api/src/app/ee is under a commercial licence, and using those enterprise features on a self-hosted instance requires an Activepieces licence.

In practice: if you install without setting an enterprise edition, you get the free Community Edition (CE): the Activepieces Docs: Self Host (Docker Compose) page states this explicitly. CE is what you self-host for free, and it is also what we run for customers.

Option 1: self-host Activepieces on a VPS

Everything below comes from the official docs as read on 5 November 2026 (Activepieces Docs: Self Host (Docker Compose), Production Setup, Activepieces Docs: HTTPS, Disaster Recovery). Activepieces moves quickly: the latest release on GitHub Releases at the time of writing was 0.92.3-hotfix.2 (28 October 2026), so check the docs again before you install.

What you need

• A Linux server with Docker Compose v2. The docs say the old docker-compose (v1) will not work.
• At least 2 vCPU and 4 GB RAM: that is the minimum the Docker Compose page lists for the recommended setup.
• Four containers: app (API and UI on port 8080), worker (runs your flows), postgres (flows, runs, connections) and redis (job queue). Redis is still part of the recommended stack.
• A domain and a TLS certificate, plus a reverse proxy in front of port 8080.

There is also a Hobbyist Installation: a single container with an embedded database (PGlite) and in-memory Redis. The docs describe it as for personal projects on one machine: it cannot scale, cannot move to Enterprise, and moving off it means starting over. Fine for trying Activepieces out; not what you want for business flows.

Install

The docs offer a one-line installer that writes the config, generates the secrets and starts everything:

curl -fsSL https://get.activepieces.com | sh
docker compose -p activepieces ps
curl http://localhost:8080/api/v1/health

If you prefer not to pipe a remote script into a shell, the same page has a manual path: clone the repository, run sh tools/deploy.sh to fill .env (it needs openssl; check that AP_ENCRYPTION_KEY, AP_JWT_SECRET and AP_POSTGRES_PASSWORD actually have values), then docker compose -p activepieces up -d. Two things the docs flag before production: the image tag is pinned, so you bump it yourself on upgrades, and the compose file runs the worker with replicas: 5, more than a small machine wants. The first account you create becomes the platform administrator.

TLS and the webhook URL

Webhooks and app triggers need an address third parties can reach. That address is AP_FRONTEND_URL in .env: in production it must be your real HTTPS domain (the docs call ngrok fine for testing but not for production). For HTTPS, Activepieces Docs: HTTPS shows an Nginx reverse proxy that redirects port 80 to 443 and proxies to localhost:8080 with the WebSocket Upgrade/Connection headers set; the certificate can come from Let's Encrypt/Certbot or Cloudflare. If the worker cannot reach the app (empty Workers page), the docs' fix is to give the worker service AP_FRONTEND_URL=http://app while the app keeps the public URL.

Backups: the part people get wrong

• Your data is not in the activepieces folder. It lives in the postgres_data Docker volume, so copying the folder does not back up your flows.
• Back up .env separately. Without AP_ENCRYPTION_KEY, stored connections cannot be decrypted, even from a full database backup.
• Postgres is the system of record; Redis is not something to back up. Disaster Recovery says not to bother backing up Redis but to enable persistence (AOF everysec), because async webhooks accepted with a 200 sit only in the Redis queue until a worker picks them up.

Updates

The documented update path is: dump the database first (docker compose -p activepieces exec postgres pg_dump -U postgres activepieces > backup.sql), read the Breaking Changes list, then run the installer with --upgrade. Your .env, data and compose edits are kept. Given how often releases ship, plan this as a recurring task, not a one-off.

Realistic effort

If you already run Docker, Nginx and Let's Encrypt on a server, the install itself is the easy part. The ongoing work is what decides it: OS and Docker security updates, Activepieces upgrades with breaking-change checks, a tested off-server Postgres backup plus a safe copy of .env, certificate renewal, monitoring that tells you when the instance or a webhook endpoint is down, and someone who is responsible when it breaks at night. If you enjoy that, self-hosting is the cheapest option. If you would rather not own it, that is what managed hosting is for.

Option 2: managed Activepieces in Germany

Full disclosure: this is what we sell. Managed Automation on overnight.host is your own isolated Activepieces (Community Edition) instance, hosted in Germany (EU), set up automatically after payment. Access details arrive by e-mail; the instance gets its own HTTPS address under <slug>.flows.overnight.host, so TLS and the public webhook URL are already in place. As listed in the shop on 5 November 2026:

This app runs in the EU region (Germany) of the overnight.host network. Our VPS, game server, voice and web plans run in the US (Dallas, TX and Charlotte, NC), with more regions planned.

Monthly billing in EUR, USD or GBP, cancel anytime, pay by card or SEPA bank transfer. All prices are final totals under the German small-business rule (§19 UStG); no VAT is shown. There is no per-task billing: what limits you is the size of the instance.

Self-host vs. managed at a glance

Short GDPR checklist

• DPA (Art. 28 GDPR) with whoever hosts the instance: your VPS provider or us.
• Connected apps are their own transfers. Hosting Activepieces does not move your CRM, mail or AI provider.
• Data minimisation and run history: pass only the fields you need, and decide how long run logs are kept.
• Record of processing (Art. 30).

The details (including what Zapier and Make themselves publish about data location) are in Zapier alternative under the GDPR and Make alternative under the GDPR.

What managed hosting does not include

• Community Edition only. Activepieces' paid enterprise features are not part of the plan.
• We don't build your flows. You get the platform; the flows are yours. Need help? Ask for a written offer by e-mail.
• It is Activepieces, not n8n. Activepieces cannot import n8n workflows; you rebuild them by hand. n8n hosting: waitlist; to run n8n yourself see n8n on a VPS.
• Support is run by one person, with no guaranteed response time.
• Not legal advice. The GDPR notes here are informational.

Migrate from n8n to Activepieces

Managed Automation runs Activepieces, an open-source workflow automation tool, as your own isolated instance hosted in Germany (EU), from €19 a month. Activepieces cannot import n8n workflows: you rebuild each workflow by hand and reconnect your accounts. Webhook URLs change, so update the services that call them.

Migrate from n8n to Activepieces → · n8n hosting: waitlist

Order Managed Automation Starter → · Managed Automation overview · Uptime monitoring (GDPR)

Written by the person who runs overnight.host: a small, honest hosting company on dedicated bare metal. Live status at up.overnight.host.

Comparing n8n options under the GDPR? Read n8n alternative under the GDPR: automation hosted in Germany: self-hosted n8n, n8n Cloud and managed Activepieces side by side, with sources.