Blog Monitoring Stack (FastAPI + MongoDB + Prometheus + Grafana + Vue.js)

This repository contains a containerized monitoring and database stack used for analyzing and visualizing the performance of a MongoDB-based FastAPI blog system. It includes MongoDB, Prometheus, Grafana, and Node/MongoDB exporters — all orchestrated via Docker Compose.

---

Quick Start

Clone the repository

git clone https://github.com/jiangfz1997/Blog_MongoDB.git
cd Blog_MongoDB

For Demo quick start:

NOTE: Make sure Docker and Docker Compose are installed on your system.

docker-compose up -d

System will be running at:

• Grafana: http://localhost:3000
• Prometheus: http://localhost:9090
• MongoDB: mongodb://localhost:27017
• FastAPI backend api doc: http://localhost:8001/docs
• Frontend: http://localhost:5173

---

For Development setup:

Set up Python environment

For FastAPI backend or script usage:

# python version 3.11+
pip install -r requirements.txt

---

Run the full system with Docker Compose

Note: Backend FastAPI server is not included in the Docker Compose setup for development purposes. It needs to be run separately (see below). Start all services in detached mode:

docker compose up -d

To stop all services:

docker compose down

If you want to clean up unused containers:

docker compose down --remove-orphans

FastAPI Development Server

Run the backend manually:

uvicorn src.main:app --reload --host 0.0.0.0 --port 8000

This backend connects to the same MongoDB instance (default URI mongodb://localhost:27017).

Swagger UI is available at: http://localhost:8000/docs

Load test

Locust is used for load testing the FastAPI backend. To run locust:

locust -f locust_test.py --host http://localhost:8000

Then open http://localhost:8089 in your browser to access the Locust web interface.

Log tracking still under development.

---

Services Overview

Service	Description	Port
MongoDB	Primary database storing blog posts, user data, etc.	27017
Mongo Exporter	Exposes MongoDB internal metrics to Prometheus.	9216
Node Exporter	Exposes host & container-level metrics (CPU, memory, disk).	9100
Prometheus	Scrapes metrics from exporters, stores time-series data.	9090
Grafana	Visualization layer for metrics dashboards.	3000

---

Grafana Dashboards

Once the containers are up:

• Visit Grafana UI → http://localhost:3000

• Default credentials: user: admin password: admin

• Add the Prometheus data source (URL: http://prometheus:9090)

• Import dashboards from the dashboards/ folder:

  • MongoDB_Instances_Overview.json – MongoDB operations and latency visualization
  • Other panels include CPU, memory, and request throughput

---

Cleanup and Maintenance

To remove all containers and volumes:

docker compose down -v

To view logs for debugging:

docker compose logs -f mongo
docker compose logs -f prometheus
docker compose logs -f grafana

Notes

• The Node Exporter installation script automatically runs inside the MongoDB container (see node-exporter-install.sh).
• If using Linux or MACOS, cadvisor can be used instead of Node Exporter for container metrics.
• You can modify prometheus.yml to add more targets or scrape intervals.
• Grafana dashboards can be customized by exporting new JSONs from the UI.

---