This project focuses on developing a comprehensive monitoring solution for microservices applications deployed on K3s. By leveraging Prometheus, Loki, Grafana, and Icinga, we aim to collect metrics, manage logs, and generate alerts efficiently.

The research scope includes:

• 📊 Evaluating monitoring requirements for microservices systems, including metric collection, log management, and alerting.
• 🛠 Deploying and configuring Prometheus and Grafana for metric collection and visualization, Loki for log management, and Icinga for system and service monitoring.
• ⚡ Conducting performance tests to assess the efficiency of the monitoring setup, comparing alert accuracy and data processing capabilities of each tool.
• 🔍 Analyzing the strengths and weaknesses of each tool in cloud infrastructure and microservices monitoring.
• 🚀 Proposing optimizations based on analysis and experimental results.

This project aims to develop an effective cloud infrastructure monitoring solution using industry-standard tools. The combination of Prometheus, Loki, Grafana, and Icinga provides a comprehensive and comparative approach to monitoring. The study not only enhances observability and incident response but also provides insights into how each toolset performs. This enables users to select the optimal monitoring solution for their system based on real-world comparisons.

The cloud infrastructure is deployed on AWS, consisting of multiple EC2 instances within a single VPC and three subnets:

• 🏗 Private subnet for K3s cluster: Hosts the K3s master and worker nodes for microservices deployment.
• 📊 Private subnet for monitoring: Hosts Prometheus, Loki, Grafana, and Icinga for system observability.
• 🌍 Public subnet:
  • ⚖️ HAProxy: Load balances traffic across K3s nodes and routes requests to appropriate backends.
  • 🔒 OpenVPN: Allows secure access to the cloud infrastructure from local environments.

To streamline deployment and configuration, the following automation tools are used:

• Terraform 🌍: Automates infrastructure provisioning.
  • Repository: terraform-hub

• Ansible 🔧: Manages server configurations.
  • Repository: ansible-hub

• Kubernetes Manifests 📦: Defines deployment configurations for K3s components.
  • Repository: kubernetes-manifests

To simulate realistic traffic and assess system performance, Locust is used to generate API requests:

• Traffic Simulation with Locust: traffic-simulation

The microservices application is available here: simple-microservices-application

A detailed evaluation of the monitoring system was conducted using the following test scenarios. The full demonstration can be found in the video:

• 🎥 Video Demo: YouTube Link

1. 🚀 Evaluating the Performance of the Load Balancing System
2. 📉 Assessing Resource Usage in the K3s Cluster
3. 🚨 Evaluating the Alerting System
   • Testing Alerts for Resource Overload
     • 🔥 API Calls to Services in the K3s Cluster
     • ⚡ Generating Load on Servers
   • Testing Alerts for Service or Server Unavailability
     • 🛑 Stopping Services in K3s
     • 🔌 Shutting Down Servers

All related documents and slides can be found in the following repository:

• 📄 Report document: Here

• 📜 Prometheus Official Documentation
• 📜 Loki Official Documentation
• 📜 Grafana Official Documentation
• 📜 Icinga Official Documentation
• 📜 K3s Official Documentation
• 📜 Terraform Official Documentation
• 📜 Ansible Official Documentation

This project provides a scalable, automated, and efficient monitoring solution for microservices on K3s. By leveraging a combination of Prometheus, Loki, Grafana, and Icinga, we establish a robust observability stack while enabling comparative analysis of these monitoring tools. The use of Terraform, Ansible, and Kubernetes Manifests ensures an automated and reproducible deployment process. 🚀