3. Installation

SciStream can be installed using three primary methods:

  1. Docker-based Installation - Recommended for most users
  2. Poetry Installation - Ideal for developers
  3. Pip Installation - For direct package installation

3.1 Prerequisites

  • Python 3.9 or higher
  • pip or poetry (recommended)
  • Docker (for container-based installation)

3.2 Docker-based Installation

SciStream is available as a Docker image, which is the simplest way to get started:

# Pull the SciStream image
docker pull castroflaviojr/scistream:1.2.1

# Create a volume for certificates
docker volume create scistream-certs

# Run a SciStream component
docker run --net=host -v scistream-certs:/scistream castroflaviojr/scistream:1.2.1 s2uc --version

3.3 Poetry Installation

Poetry provides dependency isolation for development:

# Install poetry if not already installed
curl -sSL https://install.python-poetry.org | python3 -

# Clone the SciStream repository
git clone https://github.com/scistream/scistream-proto.git
cd scistream-proto

# Install using poetry
poetry install

# Alternatively, build the package and install with pip
poetry build
pip install dist/*.whl

3.4 Pip Installation

You can also install SciStream directly using pip:

pip install scistream-proto

3.5 Verifying Installation

Verify that SciStream is correctly installed:

s2uc --version

3.6 Installing Data Plane Server Components

For a complete SciStream deployment, you'll need data plane components that handle the actual data transfer. The subprocess implementation spins up these components in an integrated way:

# Install required proxy software
# For HAProxy implementation
sudo apt-get install haproxy

# For Nginx implementation
sudo apt-get install nginx

# For Stunnel implementation
sudo apt-get install stunnel4

3.7 Next Steps

After installation, proceed to the Quickstart Guide to learn how to use SciStream.