System Requirements
For correct installation and stable operation of InsightStream RAG, you need to prepare a server with appropriate hardware and software configuration. Recommended requirements are listed below.
Hardware Requirements
Recommended server configuration:
- Linux server (Ubuntu 20.04+ recommended)
- 2 GPUs (graphics cards) NVIDIA A100 40GB
- 8 CPU threads
- 160 GB RAM
- 1 TB disk space
When scaling the system, additional resource increases may be required.
Software Environment Requirements
1. CUDA Drivers
InsightStream RAG uses GPU, so installation of the latest compatible NVIDIA drivers is required.
The preferred installation method is installation from the official NVIDIA repository. The installation process is described in detail in the documentation.
If installation from the official repository is not possible, you can use packages from the repositories of the distribution you're using.
CUDA driver installation (for Ubuntu):
sudo apt update
sudo apt install software-properties-common -y
sudo apt install ubuntu-drivers-common -y
sudo ubuntu-drivers autoinstall
sudo apt install nvidia-cuda-toolkit
2. Docker
InsightStream is deployed using Docker Compose. Make sure you have the latest version of Docker installed that supports Docker Compose V2.
Docker installation instructions for Ubuntu: 👉 https://docs.docker.com/engine/install/ubuntu/
Docker Compose V2 migration instructions: 👉 https://docs.docker.com/compose/migrate/
3. Nvidia Container Toolkit
For correct GPU operation inside containers, you need to install Nvidia Container Toolkit.
Installation instructions: 👉 https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/latest/install-guide.html
Example Environment Check Before Installation
# Check GPU availability
nvidia-smi
# Check Docker version
docker --version
# Check Docker Compose version
docker compose version
For correct operation of InsightStream, port 8080/tcp must be open for web interface access.
What's Next?
After preparing the server and software environment, you can proceed to the next stage — Installing and Deploying InsightStream.