Deploy Bash is a lightweight, ultra fast, interactive, and extensible Bash script designed to automate code deployments across multiple servers. It is built for Debian-based systems and maintained by Buffer Punk.

deploy.sh simplifies deployment workflows by handling tasks like:

• Packaging and deploying your project to multiple remote servers
• Running remote setup commands (on first deploy or updates)
• Managing symlinks for version control (via current → latest version)
• Supporting both domain and IP-based deployments
• Handling rollbacks to previous versions
• Restarting specified systemd services
• Interactive countdowns and safety prompts

The goal: make deployment effortless, safe, and repeatable — without requiring heavy CI/CD tooling.

✅ Deploy to one or multiple servers at once
✅ Supports both IPs and domain names
✅ Optional pre-deploy and setup commands
✅ Built-in rollback mechanism
✅ NPM install support for node projects
✅ apt update support (so you don't manually run it)
✅ Keeps only a specified number of releases (--keep)
✅ Validates configuration and inputs
✅ Interactive countdown before deploy
✅ Optionally adds itself to $PATH for one-line use (deploy [options])
✅ Uses human-readable color-coded outputs

• Debian-based OS (Ubuntu, Debian, etc.)
• ssh access to target servers
• bash, sudo, and scp available
• dnsutils (for domain validation)
• Remote servers must support systemctl

git clone https://github.com/bufferpunk/deploy-bash.git
cd deploy-bash
chmod +x deploy.sh
./deploy.sh

The first run will ask if you want to add it to your $PATH:

Do you want to add it to your $PATH? (y/n)

If you choose “y”, it’ll create a symlink at /usr/local/bin/deploy, so you can run it globally:

deploy [options]

You can deploy with flags or a configuration file.

deploy \
  --project=myapp \
  --type=ip \
  --servers=[192.168.1.2,192.168.1.3] \
  --services=[nginx,myapp.service] \
  --deploy-dir=/var/www \
  --setup=full
  --apt-update

Create a config file (e.g. prod.conf):

SERVERS=[192.168.1.2,192.168.1.3]
SERVICES=[nginx,myapp.service]
TYPE="ip"
PROJECT_NAME="myapp"
DEPLOY_DIR="/var/www"
SSH_USER="ubuntu"
SSH_KEY="~/.ssh/id_rsa"
NODE_HOME="." # use relative paths
SETUP_COMMAND="echo something"
# etc ...

Then deploy with:

deploy --config=deployment/prod.conf

If something goes wrong, rollback to an earlier version:

deploy --rollback=2 --config=prod.conf

This restores the second-most recent version and restarts your services.

The --setup flag allows running arbitrary setup commands on remote servers.

• --setup=full: deploys and then runs setup
• --setup=only: runs setup without deploying

You’ll be asked interactively how many setup commands you want to add, for example:

How many commands do you want to run on the remote server(s)?: 2
Enter command 1: apt update -y
Enter command 2: systemctl restart nginx

The script automatically archives deployed versions and keeps a maximum number of them (default 5). Use --keep=<number> to change that limit.

deploy --config=deployment/deploy.conf --npm

Output (simplified):

myapp/
├── deployment/
|   ├── awesome.conf
|   ├── prod.conf
|   └── your.other.scripts
├── versions/
│   ├── myapp20251020142300/
│   ├── myapp20251019115022/
│   └── ...
└── your_other_code

MIT License © 2025 Buffer Punk

Pull requests are welcome! Please open an issue first to discuss proposed changes.