
Production-ready MLOps platform for Kubernetes with:
- Apache Airflow
- MLflow
- MinIO
- PostgreSQL
- Redis
- Spark
- JupyterHub
- Prometheus
- Grafana
- NGINX Ingress
Features
- Kubernetes-native architecture
- MLflow tracking server
- Airflow workflow orchestration
- Spark job execution
- Object storage using MinIO
- Monitoring with Prometheus and Grafana
- Persistent storage
- Horizontal scaling
- Ingress-based routing
- Secrets management
- Production-ready manifests
Stack
- Kubernetes
- Docker
- Apache Airflow
- MLflow
- Apache Spark
- PostgreSQL
- Redis
- MinIO
- JupyterHub
- Grafana
- Prometheus
Deploy
kubectl apply -f kubernetes/
Namespace
Components
- airflow-webserver
- airflow-scheduler
- mlflow
- spark-master
- spark-workers
- postgres
- redis
- minio
- grafana
- prometheus
Notes
Update passwords, ingress domains, and storage classes before production deployment.
Project Roadmap
GitHub Actions
This repository includes:
- Shell validation
- Markdown linting
- Terraform validation (where applicable)
Example Deployments
See:
This repository is part of the Nexoryx infrastructure ecosystem.