TLab — Devops SRE Project

Context

For learn best pratics and developpe my competence in Devops and SRE field, i choice to learn new tools and technologies. And what is the best way to learn new tools and technologies than build a project with them.

The project is compose to 2 endpoint :

First is /health to check if project is running and this endpoint collect metrics for Prometheus.
Second is /compute to execute a load math calculation and this endpoint collect metrics for Prometheus too.

The objective is to charge with a script python to call an enpoint and see metrics in real time in prometheus and grafana.

Objectives

The first objective of project was to learn in first and in second to create an fonctional project. The second objective was to explore new tools and technologies in the DevOps and SRE field like Grafana, Prometheus, Locust, and update my knowledge in Python and FastAPI. The thrid objective was to update basique tool with AI and generate repport localy and not use extern provider of AI. The forth objective was to explore more with Gitlab CI and create artifacts.

Tech Stack

Python - Programming language
Pytest - Testing framework
FastAPI - Web framework
Docker - Containerization
Ollama - Local AI - Phi3.5 model using
Locust - Load testing
Gitlab - Source control - CI
Grafana - Monitoring and visualization
Prometheus - Monitoring and visualization

Architecture

Everything is Dockerzed and use Docker Compose to orchestrate the project, the project is composed of 4 containers :

FastAPI - Is using to expose 2 endpoind :
- /health - Check if the project is running.
- /compute - Execute a load math calculation.
Ollama - Local AI - Phi3.5 model using to generate repport.
Grafana - Monitoring and visualization
Prometheus - Monitoring and visualization
Gitlab CI - Using to generate artifacts and execute all tests.

What I Learned