Komodo

Komodo is an open-source, self-hostable platform for managing servers, builds, deployments, and automated procedures — a self-hosted alternative to managed PaaS tools like Heroku, Vercel, and Railway. The key difference is that instead of deploying to a vendor's cloud, Komodo connects to your own servers and deploys there. There is no limit on the number of servers you can connect, no usage quotas, and no "business edition" — the README states the only limit is the one you set. You keep your deployment platform on infrastructure you control.

Komodo is built from two components. Core is the web server hosting the API and browser UI that all user interaction flows through. Periphery is a small, stateless agent installed on each connected server that Core calls over a bi-directional WebSocket connection to run builds, deploy containers, report system usage, and stream container logs. Installing Periphery on a new machine is a three-step process: create an onboarding key in the UI or API, install the Periphery agent with that key, and confirm the connection status. Authentication between Core and Periphery uses automatically generated public/private key pairs, with automatic key rotation, so there's no manual key management.

Komodo supports several deployment workflows. You can deploy Docker containers — create, start, stop, and redeploy them from the UI, view status and logs, and exec into shells. You can deploy compose stacks, defining the file in the UI, on the host, or in a git repo with auto-deploy on push. It also manages Docker Swarms, so you can connect swarm managers and deploy services and stacks across a cluster. Container images are built by a dedicated Builder, and Komodo can even use AWS EC2 spot instances for scalable build capacity. Docker is the primary container engine, with Podman supported via the podman → docker alias.

Komodo Core runs as a Docker container and only needs Docker to get started — the recommended path is a Docker Compose setup backed by MongoDB, which stores resource configuration, user accounts, audit logs, and system state. Every server you want to manage additionally needs the Periphery agent, which you can install as a systemd service, a Docker container, or a raw binary. Komodo can back up its database on a schedule (gzip-compressed, stored on disk or a remote server, keeping the 14 most recent by default) and restore from any snapshot, so disaster recovery is built in. Additional configuration — including OAuth/OIDC providers like Authentik or Keycloak — is documented in the advanced setup guide.

Yes. Komodo orchestrates multi-step workflows with Procedures and Actions, which can be scheduled to run regularly — useful for backups, maintenance, and repetitive ops. Deployments can be triggered automatically by git pushes when a compose stack or build points at a git repo. It also manages configuration through shared variables and secrets with interpolation across all resources, so credentials live in one place and get injected where needed. Every action is recorded in the audit logs, and Komodo exposes a full REST and WebSocket API (with client libraries) for programmatic control, so you can build your own automation on top.