Simple Flask Application

This example demonstrates how to set up a simple Flask application with five endpoints. As you generate traffic to these endpoints, you can view the output from the /http_data endpoint and monitor metrics through Prometheus.

Deployment Instructions

Follow these steps to run the application on Minikube:

1. Install Minikube and kubectx

• Download Minikube: Visit the Minikube Installation Guide to get Minikube for your environment.
• Install kubectx: This tool simplifies context and namespace switching in Kubernetes. You can find it on GitHub - kubectx.

2. Start Minikube

Run the following command to start Minikube:

minikube start

Build or Use a Docker Image

You can either build your own Docker image or use a prebuilt one. To build an image directly in Minikube, execute:

# Configure your shell to use Minikube's Docker daemon
eval $(minikube docker-env)
cd app
docker build -t custom-prometheus-flask-exporter .

Deploy Prometheus with NGINX

Run the following commands to set up Prometheus and NGINX:

# Switch to the Minikube Kubernetes context
kubectx minikube

# Create a monitoring namespace
kubectl create namespace monitoring && kubens $_

# Generate htpasswd for authentication
htpasswd -c auth admin

# Create a Kubernetes secret for NGINX authentication
kubectl create secret generic prometheus-nginx-htpasswd --from-file=auth -n monitoring

# Add Helm repositories for Prometheus and NGINX
helm repo add prometheus-community https://prometheus-community.github.io/helm-charts
helm repo add ingress-nginx https://kubernetes.github.io/ingress-nginx
helm repo update
helm dependency update prometheus # due to Chart.lock

# Deploy Prometheus using Helm
helm upgrade --install prometheus prometheus --values prometheus/memory-alerts.yaml --values prometheus/values/values.yaml -n monitoring

# Access Prometheus via NGINX
kubectl get pods -n monitoring
NGINX="prometheus-ingress-nginx-controller-fd5c64db8-c4z6f" # Example pod name
kubectl port-forward $NGINX 8080:80 &
# Open a browser and navigate to:
http://localhost:8080 # prometheus
http://localhost:8080/alertmanager # alertmanager

Deploy the Flask Application

Now, set up the Flask application:

Note: If you prefer to use a prebuilt image, set the image to sg110/custom-prometheus-flask-exporter and comment out the line imagePullPolicy: Never in the app.yaml file.

# Create a Flask namespace
kubectl create namespace flask && kubens $_

# Deploy the Flask app (modify the image if using a prebuilt one)
helm upgrade --install flask flask -n flask

# Access the Flask application
kubectl get pods -n flask
FLASK="prometheus-flask-exporter-f6956d6b5-ww5xf" # Example pod name
kubectl port-forward $FLASK 5000:5000 &
# Open a browser and access the following endpoints:
# http://localhost:5000/index
# http://localhost:5000/action
# http://localhost:5000/error_endpoint
# http://localhost:5000/client_error_endpoint
# http://localhost:5000/http_data
# http://localhost:5000/metrics

# To send requests to each endpoint, run:
chmod +x load-test.sh
./load-test.sh

View Custom Metrics in Prometheus

In Prometheus, you can monitor the following custom metrics:

http_requests_total
http_requests_400_total
http_requests_500_total