Medusa 2 on your own VMs, Postgres next door

Deploy

A Medusa 2 store with the standard CLI

Two VMs, one config block and the standard Medusa CLI. Add the storefront VM the same way as the Next.js page describes.

~/acme-store

$ edge compute create --name medusa \

--size s-4vcpu-8gb --image ubuntu-24 \

--region london --script nodejs

✓ vm-2b7d90 running at 64.34.86.57

$ npx medusa db:migrate

$ npx medusa build

✓ Server and admin built to .medusa/server

$ cd .medusa/server && npm install

$ pm2 start npm --name medusa -- start

✓ Listening on :9000, admin at /app

Platform fee per order$0

api.acme.comRunning

Medusav2 server and worker

Postgresmedusa-db (private)

1. 01

   Create the database and Medusa VMs

   The postgresql script installs PostgreSQL 16 with remote access. Attach both VMs to one private network, then allow 5432 from that range only.

   shellCopy

   $ edge compute create --name medusa-db --size s-2vcpu-4gb \
       --image ubuntu-24 --region london --script postgresql \
       --script-param db_name=medusa
   
   $ edge compute create --name medusa --size s-4vcpu-8gb \
       --image ubuntu-24 --region london --script nodejs

2. 02

   Point the file module at Edge Storage

   Medusa 2 ships an S3 file provider. Path-style addressing and the Edge endpoint are the only changes from AWS.

   medusa-config.tsCopy

   modules: [{
     resolve: "@medusajs/medusa/file",
     options: {
       providers: [{
         resolve: "@medusajs/medusa/file-s3",
         id: "s3",
         options: {
           bucket: "medusa-media",
           region: "us-east-1",
           endpoint: "https://storage.edge.network",
           file_url: "https://media.acme.com",
           access_key_id: process.env.S3_ACCESS_KEY_ID,
           secret_access_key: process.env.S3_SECRET_ACCESS_KEY,
           additional_client_config: { forcePathStyle: true },
         },
       }],
     },
   }],

3. 03

   Migrate, build and start

   Redis lives on the Medusa VM and listens on localhost. The production server runs from .medusa/server under PM2.

   shellCopy

   $ apt install -y redis-server
   $ echo "DATABASE_URL=postgres://<user>:<pw>@10.185.0.2/medusa" >> .env
   $ echo "REDIS_URL=redis://localhost:6379" >> .env
   
   $ npx medusa db:migrate && npx medusa build
   $ cp .env .medusa/server/.env.production
   $ cd .medusa/server && npm install
   $ NODE_ENV=production pm2 start npm --name medusa -- start

4. 04

   Front the storefront and media with the CDN

   Product images then resize from the URL: media.acme.com/tee.jpg?width=800&format=auto.

   shellCopy

   $ edge cdn create --name acme-store
   $ edge cdn domains add cdn-a1b2c3 --domain shop.acme.com \
       --origin https://<storefront-ip>
   $ edge cdn domains add cdn-a1b2c3 --domain media.acme.com \
       --origin-type storage --storage-bucket medusa-media --jit-image

Prefer to hand it off? Give the job to your AI agent or have our engineers do it.