DevOps Project: “Weather & Alerts Microservices Platform”

Project Overview

The idea is to build a CI/CD for Python-based microservices platform that fetches weather data and generates alerts. The application should be containerized, deployed to a local Kubernetes cluster (Minikube), monitored using Prometheus and Grafana, and delivered via an automated Jenkins CI/CD pipeline.

---

Problem Statement

Goal:

Design, build, deploy, monitor, and automate a multi-container Python microservices platform using DevOps tools.

---

Requirements

1. Application

Provided a simple Python application to fetch the weather information from OpenWeatherMap API three Python Flask microservices: Download the app and use your github personal account to do the exercise and share the repo url as end outcome.

Service Name	Port	Role
gateway	4000	Entry point, routes API calls
weather-service	5000	Fetches live weather data (OpenWeatherMap API)
alert-service	6000	Generates alerts based on weather data

• Each service should expose /health and /metrics endpoints.
• All APIs return JSON responses.
• All services should be containerized using Docker.

---

2. Dockerization

• Create a Dockerfile for each service.
• Images must be tagged properly (gateway:latest, weather-service:latest, etc.)
• Containers must run using port as defined above.

---

3. Kubernetes Deployment

Deploy each microservice to Kubernetes (Minikube) using raw YAML manifests.

• Create a Deployment and Service for each microservice.
• Expose gateway via NodePort so it’s accessible on your browser.
• Make sure services can talk to each other using service DNS (e.g., http://alert-service:6000/alert).

---

4. CI/CD with Jenkins

Set up a Jenkins pipeline that automates the following:

• Checkout code from your local GitHub repo
• Build Docker images for all services
• Deploy services to Kubernetes using kubectl apply
• Include basic health checks using curl

---

5. Monitoring with Prometheus + Grafana

• Integrate prometheus_flask_exporter into each service

• Each service must expose /metrics

• Deploy Prometheus to Kubernetes to scrape metrics

• Deploy Grafana and create dashboards to visualize:

  • API response time
  • Alert count
  • Error rate
  • CPU/memory usage (node-exporter)

---

6. Infrastructure Setup

Provision your laptop with the necessary tools using:

• Ansible to install:

  • Docker
  • Minikube
  • Kubectl
  • Jenkins

---