AWS EC2 Grafana

Configure and deploy a Grafana server on AWS EC2 to display a public dashboard with PostgreSQL data.

This project uses Grafana, Nginx, Docker and AWS CLI. All steps are performed using the AWS CLI and shell scripts.

Starting point: an AWS account with:

• An EC2 instance with access to a VPC and with internet access
• A PostgreSQL database on the VPC, without internet access (e.g. hosted on AWS RDS)

View the example deployment at: https://grafana.e-mit.dev

See also

• github.com/e-mit/aws-create-db for the EC2 and RDS setup
• github.com/e-mit/aws-lambda-get and github.com/e-mit/aws-lambda-db for gathering API data into the RDS using Lambda functions

Readme Contents

• Overview
  
• Development and testing
  
• Release build and deployment
  

Overview

• The project builds two Docker containers: Nginx and Grafana.
• The Grafana container provides a public dashboard showing graphs for data in either an in-container SQLite database (for testing), or a remote PostgreSQL database.
• The Nginx container is configured as a reverse proxy, rate limiter and TLS terminator.

Development and testing

Build and run locally (without AWS interaction), choosing either SQLite or PostgreSQL as data source, using:

• grafana_local_test.sh for the Grafana container without Nginx
• compose-local-test.sh for Grafana and Nginx (without TLS)

GitHub Action workflows build, run and test the local SQLite configuration on every push.

Release build and deployment

1. Build and push to Dockerhub using release-build.sh
2. Do a one-time preparation of the EC2 instance using ec2_setup.sh - this does the following:
   • Install packages
   • Create an AWS Security Group with inbound TCP permissions
   • Obtain TLS certificates from Let's Encrypt
   • Setup periodic jobs for certificate renewal and security update installation
3. Deploy the project to the EC2 using ec2_deploy.sh

Steps 1 and 3 can be repeated after changes; step 2 is unaffected and does not require repeating.