Skip to content

Run your apps on your own servers.

Deploy from Git and get databases, monitoring, encrypted backups and failover when a server goes down.

Free for personal use · Enterprise from $9 per server a month

Failover test, replayed
node-1 10.0.0.11 node-2 10.0.0.12 node-3 10.0.0.13 your app shop.example.com
Three servers share the work. Your app is served from all of them.
  1. all 3 servers healthy

Latest release v1.2.0 Changelog Open roadmap and issues GitHub

  • Failover tested by powering a server off while requests keep coming. The test

  • Encrypted nightly backups, with a restore test that checks them. Backups

  • Two-factor sign-in, roles per project and an audit log of changes. Sign-in

  • A firewall and automatic security updates on every server. Hardening

See how it works. Short recordings of a real install.

Each clip is a screen recording of Cairn on real servers, not an animation. Long waits are shortened, and the clip says so.

1Install

One command sets up a server. Kubernetes and everything on top.

Run platformctl provision on an Ubuntu server. It checks your backup bucket first, then installs Kubernetes, the portal, databases and monitoring, and prints where to sign in.

How the install works

2Deploy

Push to Git, see it live. Every deployment tracked.

Link a repository and pick a branch for each environment. Cairn deploys the Kubernetes files in it and shows each deployment, its commit and its health on the project page.

How it works

  1. git pushto your branch
  2. Your CIbuilds the image
  3. Cairnsyncs k8s/
  4. Liveover HTTPS
How deploys work

3Observe

Logs, metrics and traces. For every project, built in.

Open a project's Monitoring section to search its logs, watch request rates and follow one slow request through your services. Traces need the full install profile and the Cairn SDK.

How monitoring works

4Back up

Backups you have restored. Not just backups you have taken.

Send encrypted backups to your own S3-compatible bucket, then run a restore test that restores the latest one into a scratch database and checks it. Settings shows when each last ran.

How it works

  1. Serversdata, settings
  2. Encryptedevery night
  3. Your bucketany S3 store
  4. Restore testchecks it
How backups work

5Stay up

Turn a server off. The others take over.

Run Cairn on three servers. When one loses power, the others restart its apps and take over its databases, with at most a few seconds of errors, and it rejoins on its own.

How it works

  1. One serverruns it all
  2. Join twomore servers
  3. Copieson each server
  4. One failsothers take over
How failover works

6AI agents

Let an AI assistant help. A person approves risky changes.

Paste one connector URL into Claude, ChatGPT, Cursor or VS Code. The assistant reads health, logs and pods; deleting, scaling or reading secrets waits for a person to approve.

How AI access works

Also included. On the same servers.

  • A database per project

    PostgreSQL, MongoDB and Redis for each environment, each with its own user.

    Databases
  • Alerts and on-call

    Alerts to email, Slack, Discord, PagerDuty or a webhook, with escalation and rotations.

    Alerts
  • Public status pages

    One page per project, fed by your uptime checks, with incidents and maintenance notices.

    Status pages
  • Errors and real users

    Errors grouped into issues by release; session replay and heatmaps with the browser SDKs.

    Error tracking
  • Preview environments

    Each branch other than main gets a temporary deployment and address.

    Previews
  • Builds from source

    Turn on builds and Cairn builds your image from a Dockerfile or with Buildpacks.

    Builds
  • Built on open-source tools

    Runs on Kubernetes
    k3s, Argo CD, ingress-nginx, cert-manager
    Databases
    PostgreSQL, MongoDB, Redis
    Storage
    MinIO, Longhorn
    Monitoring
    Prometheus, Grafana, Loki, Tempo, Pyroscope
    Every component and version

How Cairn compares. Self-hosted tools and hosted platforms.

Cairn Coolify CapRover Dokku Heroku, Render
Runs on your own servers Yes Yes Yes Yes No
Several servers, with database failover Yes Partial Partial — Paid plans
Metrics, logs and traces built in Yes Partial Partial Logs Partial
Encrypted off-site backups, with a restore test Yes Partial — Via plugins —
AI assistant access, with a person approving risky actions Yes — — — —
Free to use Personal use Yes Yes Yes Render, limited

Based on public docs, November 2026; tell us if something changed. “Partial” means some of it, or only with extra setup; a dash means we could not confirm it in their docs.

Your servers, your data. Start with one Ubuntu server.

On a fresh Ubuntu server
$ curl -fsSL https://get.cairnpaas.dev | bash
$ sudo platformctl provision

It asks for your domain, an admin email and where backups go, then prints the address to sign in.