Quick Start#

This guide starts the Scenescape demo with one command. For more configurable deployment options, see the installation.

Prerequisites#

  • Review the System Requirements.

  • Clone or extract the Scenescape repository and change to its root directory.

  • Run as a standard, non-root user who owns the repository checkout and can use sudo; the script uses sudo to install missing host packages and tools.

  • Have network access for package, image, model, and tool downloads.

Installation#

From the repository root, run the bootstrap script:

./deploy.sh

To preview the dynamic pipeline configuration for cameras, run the bootstrap script in Kubernetes mode:

KUBERNETES=1 ./deploy.sh

What deploy.sh Does Automatically#

The script checks that it is running on a supported Debian/Ubuntu or Fedora/RHEL system, that sudo is available, and that the checkout is owned by a standard non-root user. It installs missing host packages with the system package manager:

  • Debian/Ubuntu: git, curl, make, openssl, and unzip.

  • Fedora/RHEL: git, curl, make, openssl, unzip, and nmap-ncat.

On Debian/Ubuntu, it also installs Docker if the Docker Compose plugin is missing. On Fedora/RHEL, Docker and Docker Compose must already be installed. The script checks installed Docker and Compose versions, generates passwords and TLS/authentication secrets, builds the core images, starts the demo video sources, uploads the demo scenes, and prints the generated SUPASS for login.

With KUBERNETES=1, the Kubernetes Makefile additionally installs missing kind, kubectl, and Helm tools (and optional k9s), creates or recreates the default Kind cluster, installs Cert Manager, loads the core images, deploys the chart, starts the video sources, and imports the demo scenes.

Some steps are conditional or interactive: the script asks for alternate ports if the default HTTPS or MQTT port is occupied, asks whether to upgrade an existing database when needed, and may use sudo to add the user to the Docker group. The Kubernetes flow deletes and recreates the default Kind cluster, so do not run it if that cluster contains workloads or data you need to keep.