Skip to main content

Quick start

Goal: a running Octostar platform on a cluster with internet access. For the full procedure, see Online installation. For sites without internet, see Air-gapped installation.

Before you start, check Prerequisites.

1. Download the installer​

curl -fsSL https://install.octostar.com/install-octostar-raw.sh | bash -s -- --nightly
sudo chown -R "$USER" /opt/octostar
cd /opt/octostar

2. Configure​

cp local-env.template.yaml local-env.yaml

Edit local-env.yaml. Set at least:

  • dockerHub.token. Do not use only the DOCKERHUB_TOKEN variable: the image pull secret reads the file.
  • timbrAdminPassword. The default password is public.
  • octostar.domain, if you do not use local.test.

⚠️ Decide before the first install — storageClass and the write-once secrets cannot change later. See Configuration.

3. Install​

./bin/install.sh --context <context-name>
  • helm, helmfile, kubectl and yq come with the release, so on Linux and on macOS with Apple silicon there is nothing to install first.
  • Add --dry-run to preview without changing the cluster.
  • Add --no-wait-downloads if you enable map data and do not want to wait for it.

The install takes a long time. It stops before it changes anything if the cluster check finds a problem.

4. Verify​

The install uses the release's own kubectl. To run it from your shell, put the release's tools on your PATH first. Run this in /opt/octostar (see Prerequisites):

export PATH="$PWD/bin/tools/$(uname -s | tr '[:upper:]' '[:lower:]')-$(uname -m | sed 's/x86_64/amd64/; s/aarch64/arm64/'):$PATH"
kubectl get pods -n octostar-main

When the pods are Running or Completed, open https://home.<domain> (default https://home.local.test). Log in as admin with timbrAdminPassword. With the bundled Keycloak, see Configuration → Admin access.

For the default local.test self-signed certificate, trust the root CA (certs/rootCA.crt) on the machine with your browser. --install-cert does this only for the host that runs install.sh. It does not stop after that step: it runs a full install. To trust the CA on that host, add --install-cert to the install command in step 3. On a working installation, the flag ends with an "already installed" error.

If it fails​

  • Select the cluster with kubectl config use-context <context-name>, then run ./bin/install.sh --dump-logs --context <context-name>. The log tool ignores --context, and uses the current context. Send the file to Octostar support.
  • The cluster check names the problem and the way out. See Shared clusters.
  • See Operations & uninstallation.

Next​

Upgrading, Backup & restore and Operations & uninstallation.