Traffic simulation

This repository contains three comprehensive scenarios for load testing, monitoring, and performance analysis of applications running on Kubernetes (K8s) with HA Proxy, Locust, and additional monitoring tools.

Setup

1. Create a virtual environment

# Install virtualenv if you don't have it already
pip install virtualenv

# Create virtualenv
virtualenv venv

# Activate the virtual environment
# On macOS & Linux
source venv/bin/activate

# On Windows
venv\Scripts\activate

2. Install Locust

pip3 install locust

1. Load Balancing Evaluation with HA Proxy

Overview

This scenario focuses on measuring the performance of the system as HA Proxy distributes load to pods within a Kubernetes cluster.

Steps

• Run Locust: Simulate a large number of virtual users sending requests to the application through HA Proxy.
• Monitoring: Use Prometheus and Icinga to monitor traffic through HA Proxy.
• Log Collection: Utilize Loki to collect logs from HA Proxy and the application pods to identify any errors during high load conditions.

Running the Tests

locust -f load_test.py --host=http://your-haproxy-host

Metrics to Evaluate

• Requests per second: Measure the number of requests handled by HA Proxy.
• Load Distribution Performance: Assess if the load is evenly distributed among the pods.
• System Latency: Monitor latency as the number of requests increases.

2. Heavy Load Testing with MySQL

Overview

In this scenario, we create a substantial dataset to test MySQL's processing capabilities under high load conditions.

Steps

• Locust Scripting: Develop a Locust script to simulate multiple concurrent users sending read/write queries to MySQL.
• Gradual Load Increase: Run Locust while gradually increasing the number of virtual users from a low to a high number.

Running the Tests

locust -f mysql_load_test.py --host=http://your-app-host

Metrics to Evaluate

• Average Response Time: Track the average response time of the application as the user count increases.
• MySQL Queries per Second: Measure the number of MySQL queries processed per second.
• Latency and Resource Utilization: Monitor MySQL's latency and resource usage under heavy load.

3. Performance Evaluation with Docker and Harbor

Overview

This scenario evaluates the performance of the system while deploying Docker images to Kubernetes through Harbor.

Steps

• Build Docker Image: Create a Docker image for the application and push it to Harbor.
• K8s Deployment Configuration: Prepare configuration files for deploying the application to Kubernetes.
• Scaling and Deployment Testing: Increase the number of replicas and monitor the time taken to deploy images from Harbor to Kubernetes.

Running the Tests

docker build -t your-harbor-repo/your-app:latest .
docker push your-harbor-repo/your-app:latest

kubectl apply -f deployment.yaml

Metrics to Evaluate

• Image Retrieval and Deployment Time: Measure the time taken to fetch and deploy the Docker image from Harbor.
• Kubernetes Resource Usage: Monitor CPU and RAM utilization in Kubernetes during the deployment process.
• Log Analysis: Analyze logs to detect and address any deployment errors.

Conclusion

This repository provides a comprehensive framework for load testing, monitoring, and performance evaluation. Each scenario is designed to give insights into the system's behavior under various load conditions, helping to ensure a robust and efficient application deployment in a Kubernetes environment.