Quick install

The installer sets everything up on a Linux server or VPS: it installs Docker if needed, logs in to the private registry with your license, downloads KumoYami's files, sets up HTTPS with Caddy if you give it a domain, starts the site and shows you the code for the first-run setup.

Before you startCopy Anchor Link

• A Linux server (Debian, Ubuntu, Rocky, Alma…) with root or sudo access, at least 1 GB of RAM and 5 GB of free disk space.
• Your license ID (lic_…) and key (KY1.…), from your purchase email.
• If you want HTTPS from the start: a domain (for example nube.example.com) with an A record pointing at the server's IP, and ports 80 and 443 free and open in the firewall.

InstallCopy Anchor Link

curl -fsSL https://kumoyami.com/install.sh | sudo bash

If you didn't give it a domain, KumoYami is only reachable from the server itself. To open the panel from your own computer, start an SSH tunnel and visit http://localhost:4890/admin:

ssh -L 4890:127.0.0.1:4890 root@SERVER-IP

When you're ready for a domain, run the installer again with --domain, or use your own reverse proxy.

What it does, step by stepCopy Anchor Link

1. Checks the system is Linux on x86_64 or aarch64, that it's running as root, and that there's enough memory and disk (it warns below 1 GB of RAM or 5 GB free).
2. Installs Docker and the docker compose plugin if they're missing, and starts the service.
3. Logs in to registry.kumoyami.com with your license (docker login). If there's already a session, it asks whether to keep it.
4. Creates the install folder, /opt/kumoyami by default, with:
   • docker-compose.yml, downloaded from kumoyami.com/install/docker-compose.yml.
   • .env, created from kumoyami.com/install/env.example (the first time only; after that it's kept).
   • datos/, where KumoYami keeps its settings, owned by the container's user (uid 1000).
5. With a domain: checks that ports 80 and 443 are free and that DNS points at the server, then writes a Caddyfile and a docker-compose.override.yml with Caddy.
6. Pulls the image, starts the containers and waits up to 2 minutes for the setup code to appear.

You can run it again whenever you like: it updates docker-compose.yml (keeping the old one as docker-compose.yml.bak-DATE), keeps .env and your data, then pulls and restarts the containers.

OptionsCopy Anchor Link

With curl | bash, options go after bash -s --:

curl -fsSL https://kumoyami.com/install.sh | sudo bash -s -- --domain nube.example.com --email you@example.com

Environment variablesCopy Anchor Link

With sudo, variables go after sudo (any you've exported don't make it through). For example, an install with no questions asked:

curl -fsSL https://kumoyami.com/install.sh | sudo KUMOYAMI_LICENSE_ID=lic_xxxxxxxxxxxx \
  KUMOYAMI_LICENSE_KEY='KY1.…' bash -s -- --unattended --domain nube.example.com

--unattended needs the ID and the key, unless the server is already logged in to the registry.

Handy commandsCopy Anchor Link

In the install folder (cd /opt/kumoyami):

docker compose logs -f web                    # follow the logs (Ctrl+C to exit)
docker compose pull && docker compose up -d   # update
docker compose logs web | grep -i código      # show the setup code again

Settings live in /opt/kumoyami/.env and data in /opt/kumoyami/datos.