acme-website

ACME Corp Marketing Website

The Story

ACME Corp is migrating their marketing website off Heroku. The site is a Node.js application backed by a PostgreSQL database that stores page content, blog posts, and lead-capture form submissions. The team wants infrastructure as code so every change is reviewed in a pull request, and they need automated daily backups shipped to S3-compatible storage so the marketing data is never at risk.

This scenario provisions the entire stack (project, database, application, environment wiring, and an optional backup schedule that targets an existing UI-managed S3 storage) in a single terraform apply.

What Gets Created

#	Resource	Purpose
1	coolify_project.acme	Logical grouping for all ACME resources
2	coolify_database_postgresql.content	PostgreSQL 16 database for page content
3	coolify_application.website	Node.js marketing site from Git
4	coolify_environment_variable.database_url	Connects the app to the database
5	coolify_environment_variable.node_env	Sets NODE_ENV=production for builds
6	coolify_database_backup.daily	Daily backup schedule using an existing UI-managed S3 storage (optional)

Architecture

┌─────────────────────────────────────────────────────┐
│                  Coolify Project                    │
│                  "acme-website"                     │
│                                                     │
│  ┌──────────────────┐    ┌───────────────────────┐  │
│  │   PostgreSQL 16   │    │   Node.js Application │  │
│  │   "acme-content"  │◄───│   "acme-website"      │  │
│  │                   │    │                       │  │
│  │   postgres:16     │    │   nixpacks build      │  │
│  │   port 5432       │    │   port 3000           │  │
│  └────────┬──────────┘    └───────────────────────┘  │
│           │                        ▲                 │
│           │              ┌─────────┴─────────┐       │
│           │              │ Environment Vars   │       │
│           │              │ DATABASE_URL ──────┼───┐   │
│           │              │ NODE_ENV=production│   │   │
│           │              └───────────────────┘   │   │
│           │                                      │   │
│           └──────────────────────────────────────┘   │
│                                                     │
└─────────────────────────────────────────────────────┘

           ┌──────────────────────────────────┐
           │ Existing UI-Managed S3 Storage    │
           │       "acme-backups"              │
           │                                   │
           │  ◄── Daily Backup ── PostgreSQL   │
           │       (@daily cron)               │
           └──────────────────────────────────┘

Prerequisites

1. Running Coolify instance: a server with Coolify installed and accessible via its API endpoint (e.g. https://coolify.example.com).
2. API token: generate one in Coolify under Security > API Tokens.
3. Server UUID: the UUID of the destination server registered in Coolify. Find it under Servers in the Coolify dashboard.
4. (Optional) Existing S3 storage UUID: only needed if you enable backups (enable_backups = true). Create the storage in the Coolify web UI first, then pass its UUID with existing_s3_storage_uuid. The storage can target AWS S3, MinIO, Backblaze B2, Cloudflare R2, or another compatible backend.

Usage

Initialize the workspace:

terraform init

Preview the changes:

terraform plan \
  -var="coolify_endpoint=https://coolify.example.com" \
  -var="coolify_token=your-api-token" \
  -var="server_uuid=your-server-uuid"

Apply:

terraform apply \
  -var="coolify_endpoint=https://coolify.example.com" \
  -var="coolify_token=your-api-token" \
  -var="server_uuid=your-server-uuid"

To enable daily backups to S3, first create the storage in the Coolify web UI, then pass its UUID here:

terraform apply \
  -var="coolify_endpoint=https://coolify.example.com" \
  -var="coolify_token=your-api-token" \
  -var="server_uuid=your-server-uuid" \
  -var="enable_backups=true" \
  -var="existing_s3_storage_uuid=your-s3-storage-uuid"

How It Works

Coolify runs every resource as a Docker container. All resources within the same project are placed on the same Docker network. This means containers can reach each other using their resource name as the hostname, so no need to expose database ports to the public internet.

In this scenario the DATABASE_URL environment variable uses the database resource's name attribute as the hostname:

postgresql://acme:PASSWORD@acme-content:5432/acme_content

acme-content resolves to the PostgreSQL container because Coolify's internal Docker networking handles DNS within the project network. The database stays private (is_public = false) while the application connects over the internal network.

Clean Up

Destroy all resources created by this scenario:

terraform destroy \
  -var="coolify_endpoint=https://coolify.example.com" \
  -var="coolify_token=your-api-token" \
  -var="server_uuid=your-server-uuid"