Skip to content

README.md

Latest commit

 

History

History
74 lines (56 loc) · 3.41 KB

README.md

File metadata and controls

74 lines (56 loc) · 3.41 KB

Deployment

The deployment directory contains necessary files for the deployment of Meet@Mensa application.

To deploy the system manually on your local machine, follow the guides below. Automated deployment workflows in GitHub Actions can be found under ../.github/workflows.

Docker Compose

Prerequisites:

  • a .env file is located in /docker directory and contains the OPENAI_API_KEY secret for the GenAI service (in AWS deployment, a GitHub secret is used instead)

Use the following commands to start / stop / inspect the project using Docker Compose:

cd docker                       # Move to docker directory
docker compose up --build       # Starts project in the foreground (with force re-build)
docker compose up -d            # Starts project in the background
docker compose ps               # Shows all running containers
docker compose down             # Stops the project
docker compose logs             # Shows the logs of all containers
docker compose pull             # Pulls the images of all services
docker images                   # Shows locally chached images

If the client container is crashing with this error 'exec /entrypoint.sh: no such file or directory', this may be caused by line-endings differing in Windows and Linux. To fix this switch between CRLF (Windows) to LF (Linux).

Kubernetes Helm

Prerequisites:

  • you are already authenticated in a running Kubernetes cluster (see Kubeconfig file),
  • the microservice images are built and pushed to GHCR (see README.md in individual microservices),
  • database access passwords already exist as secrets in the devoops namespace:
kubectl create secret generic <match/user>-db-secret \
  --from-literal=MYSQL_ROOT_PASSWORD='<your-password>' \
  --namespace devoops
  • OpenAI API key is saved as secret in the devoops namespace:
kubectl create secret generic genai-secret \
  --from-literal=OPENAI_API_KEY=sk-... \
  --namespace devoops

Once the three requirements are fulfilled, you can run the application:

helm install meetatmensa ./k8s -n devoops   # Installs the Helm chart and deploys it to devoops namespace
kubectl get all -n devoops                  # Shows all Kubernetes resources deployed in devoops namespace
helm uninstall meetatmensa -n devoops       # Deletes all resources deployed with the Helm chart

If you want to access cluster-internal services, you can utilize port-forwarding: kubectl port-forward svc/meetatmensa-<service> <port>:80 -n devoops. Do not use ports 8080, 8081 or 8082 - they are reserved as entry points to the application (client service).

AWS Deployment

For AWS deployment, use the compose.aws.yml file which is configured for production deployment with SSL certificates and monitoring.

Access Points

📊 Monitoring

Meet@Mensa includes a comprehensive monitoring stack with Prometheus for metrics collection, Grafana for visualization, and Loki for log aggregation.

Local Deployment (Docker Compose)

When running locally with Docker Compose, monitoring services are available at:

  • Grafana Dashboard: http://localhost:3000

    • Default credentials: admin / admin
    • Access dashboards for application metrics, logs, and system health

  • Prometheus: http://localhost:9090

    • View metrics, alerts, and targets
    • Access alert rules and their current state