Deployment¶
Three ways to run Quantum Pipeline: install the PyPI package directly, run a single Docker container, or bring up the full platform with Docker Compose.
-
Docker Basics
Container images (CPU, GPU, Spark, Airflow), building from source, and running individual containers.
-
Docker Compose
Full platform deployment with Kafka, Spark, Airflow, Garage, MLflow, and monitoring exporters.
-
GPU Acceleration
NVIDIA setup, CUDA build args, Qiskit Aer GPU backend, and performance benchmarks from the thesis experiments.
-
Environment Variables
Reference for all
.envvariables used across services in the Docker Compose deployment.