How to move a project from your laptop to the cloud

Steps

1. Make sure secrets are not in the repository

   Run this on the laptop, in the project. .env must be ignored, and git must not track it. If the second command prints a file, it is already tracked: take it out of git, and treat every key in it as leaked.

   $ git check-ignore -v .env
   $ git ls-files | grep -i env

2. Commit and push the code

   Everything that is not a secret and not a generated file goes through git. Keep node_modules, .venv and build output out of it.

   $ git status
   $ git add -A && git commit -m "ready to move"
   $ git push

3. Clone the project on the server

   Log in as a normal user, not root. For a private repository, create a key on the server and add the public half to the repository as a read-only deploy key (on GitHub: Settings, Deploy keys).

   $ ssh [email protected]
   $ sudo apt install -y git
   $ ssh-keygen -t ed25519 -C "server"
   $ cat ~/.ssh/id_ed25519.pub
   $ git clone [email protected]:you/your-project.git && cd your-project

4. Copy the secrets by hand

   Run this on the laptop. The file travels over SSH and lands readable only by its owner. Do not paste keys into chats or tickets.

   $ scp .env [email protected]:~/your-project/.env
   $ ssh [email protected] 'chmod 600 ~/your-project/.env'

5. Copy the data git does not hold

   Uploads, local files and databases come next. Add -n for a dry run first: it lists what would be copied and changes nothing. A slash at the end of the source copies the contents of the folder, not the folder itself.

   $ rsync -avzn --progress ./uploads/ [email protected]:~/your-project/uploads/
   $ rsync -avz --progress ./uploads/ [email protected]:~/your-project/uploads/

   A Postgres database moves as a dump. Create an empty database with the same name on the server first.

   $ pg_dump -Fc mydb > mydb.dump
   $ scp mydb.dump [email protected]:~
   $ ssh [email protected] 'pg_restore -d mydb ~/mydb.dump'

6. Install the dependencies from the lockfile

   Use the lockfile, so the server gets the same versions as your laptop. Check that the runtime matches too: compare node -v or python3 --version on both machines.

   # Node
   $ npm ci
   # Python
   $ sudo apt install -y python3-venv
   $ python3 -m venv .venv && . .venv/bin/activate
   $ pip install -r requirements.txt

7. Start the app and open its port

   On your laptop the app listens on localhost. On a server it must listen on 0.0.0.0, or only the server itself can reach it. Start it, look at the listening sockets, then open the port in the firewall. Allow SSH first, or you lock yourself out.

   $ npm start             # the app must listen on 0.0.0.0:3000
   $ ss -tlnp | grep 3000
   $ sudo ufw allow OpenSSH
   $ sudo ufw allow 3000/tcp
   $ sudo ufw enable

   To keep it running after you close SSH, put it in tmux or a service: the next guide shows both.

Check that it works

Ask for the page from the laptop, not from the server. If this prints HTTP/1.1 200, the project is live at the server's address.

$ curl -I http://203.0.113.10:3000

Then close the laptop and request it again. It still answers, because the app now runs on the server.

What this guide does not cover

• It moves a project to one server. It sets up no load balancer, no second copy and no staging.
• A reverse proxy and an https certificate are a separate step. Without them the app answers on plain http and a port number.
• MySQL, MongoDB and others have their own dump tools. Only Postgres is shown here.

Common problems