Skip to content

Latest commit

 

History

History
50 lines (36 loc) · 1.47 KB

overview.md

File metadata and controls

50 lines (36 loc) · 1.47 KB

Deployment Guide

This guide describes how to build, run, and configure the NEDC Bench API in Docker Compose and Kubernetes.

Prerequisites

  • Docker and Docker Compose
  • Optional: Kubernetes cluster + kubectl

Environment Variables

  • NEDC_NFC (default: /app/nedc_eeg_eval/v6.0.0)
  • REDIS_URL (default: redis://redis:6379 in Compose; cluster DNS in K8s)
  • LOG_LEVEL (default: info; controls uvicorn and app logging)
  • MAX_WORKERS (default: 1; entrypoint sets uvicorn --workers)

Build and Run (Compose)

docker-compose build
docker-compose up -d
docker-compose ps
curl http://localhost:8000/api/v1/health
curl http://localhost:8000/metrics

Kubernetes Deployment

kubectl apply -f k8s/
kubectl get deploy,svc
kubectl rollout status deploy/nedc-bench-api
kubectl port-forward svc/nedc-bench-api 8000:80
curl http://localhost:8000/api/v1/health

Note: In production, deploy Redis, Prometheus, and Grafana using their official Helm charts. This repo intentionally includes only API manifests; configure ServiceMonitors and dashboards in your platform repo.

Image Build and Tagging

# API image
docker build -f Dockerfile.api -t nedc-bench/api:latest .
docker run --rm -p 8000:8000 nedc-bench/api:latest

Configuration Matrix

  • Compose: REDIS_URL=redis://redis:6379; Prometheus scrapes api:8000 via monitoring/prometheus.yml.
  • K8s: use cluster DNS (e.g., redis.default.svc.cluster.local:6379) if Redis runs in‑cluster.