App Deployment on Digital Ocean
Set up a CentOS server on DigitalOcean for Node.js deployment
This document outlines one method of setting up a Node.js-ready web server running CentOS, hosted by DigitalOcean.
Disclaimers:
- There are certainly other ways to deploy a Node.js app on a DigitalOcean Droplet which will work just as successfully.
- The steps outlined below work at the time of this writing, but may have to be adjusted to match changes in the technologies used as time goes by.
- The link to Digital Ocean above includes a referral code - this will give you $200 free credit. Considering that a simple Digital Ocean Droplet server costs around $5/month, this credit will serve you very well. I will receive a discount on my own account should you begin paying for DigitalOcean once your free credit runs out.
Create a Digital Ocean Droplet
- Create an account on Digital Ocean
- Create a Droplet - select the cheapest plan with no extra options
- Choose CentOS as the operating system
Create a local SSH key
Create an assymetric pair of keys locally (on your own computer):
ssh-keygen -t rsa
Save the keys into the default file presented. Do not set a password.
This creates two keys in two files:
- id_rsa - your private key… never share this with anyone
- id_rsa.pub - your public key… feel free to share this
Copy the SSH key to Digital Ocean
Locate the file with the public key you created in the previous step and copy the contents to the clipboard.
The following example shows how to do this for a user named foobar on a Mac.
cat /Users/foobar/.ssh/id_rsa.pub | pbcopy
Now paste this into the Digital Ocean SSH Key page and save.
Log into the Digital Ocean Droplet
With SSH keys in place on both your local machine and the Digital Ocean server, you can log in:
ssh root@123.456.789
Note: if you receive a Permission denied error when running ssh, you can try specifying the location of the private key file with the -i flag, e.g. ssh -i ~/.ssh/id_rsa root@123.456.789.
Update the system
Update any out of date CentOS packages:
yum -y update
Install a variety of development tools:
yum -y groupinstall "Development Tools"
A basic text editor, vim:
yum -y install vim
And a better text editor, emacs:
yum -y install emacs
Networking tools:
yum -y install net-tools
Git version control:
yum -y install git
Install Node.js
Downloading of Node.js packages is handled on CentOS by running a bash script. The following command downloads the bash script and pipes it to bash for execution:
curl -sL https://rpm.nodesource.com/setup_12.x | bash -
Node.js is not yet installed. To install it, run the following: on CentOS
yum install -y nodejs
Now Node.js is installed. Trust but verify. Run the following to check the versions of node and npm:
node -v
npm -v
Install a few common modules globally
Install the popular Node.js app manager, PM2. This is used to keep a Node.js app running as a daemon and restarts it automatically if it crashes:
npm install -g pm2
If you desire PM2 to launch automatically when the server boots up and start any Node.js applications it has been set to manage, the following command will output a command which you can copy/paste and run:
pm2 startup systemd
Install simple http server so we can quickly serve web pages for testing:
npm install -g http-server
Try serving a simple web page
Create a super-simple home page in a file named index.html:
echo "Hello World!" > index.html
Run the http-server module:
http-server
Various IP addresses of the server should be displayed. The address 127.0.0.1 is a private address that only refers to the server from the server itself - ignore it. Copy the next one, which is your IP address on the Internet.
Paste the IP address into the address bar of a web browser and hit Enter.
The web page should show up in the web browser.
If so, congrats - you have set up your own web server!
You may stop this server and delete the index.html file whenever you desire.
Create another user
It is considered a bad idea to do any development as the root superuser.
Create a regular user account that you will use to set up an app:
adduser foobar
Add this user to the group of administrators of this machine. On CentOS, all users in the wheel group can use the sudo command to run commands as the root administrator user:
usermod -aG wheel foobar
Set a password for this user. Enter a password when prompted:
passwd fooobar
Switch to regular user account
Switch to the new user account:
sudo su - foobar
whoami
Change to this user’s home directory:
cd ~
Set up the SSH keys to allow connections without passwords to this account. Only the user itself should be able to see this directory:
mkdir .ssh
chmod 700 .ssh
Create a file inside that directory:
touch .ssh/authorized_keys
chmod 600 .ssh/authorized_keys
Copy and paste the same SSH public key from your local machine that you pasted into Digital Ocean’s SSH key page into this authorized_keys file using a command-line text editor of your choice, such as vim or emacs.
Save the file and quit the text editor.
Log out of the server
First return from the regular user account back to the root account:
exit
Log out of the server completely:
logout
Log in as a regular user
SSH again to the server as the regular user account. Replace the example username and IP address below with your own:
ssh foobar@123.456.789
Note: if you receive a Permission denied error when running ssh, you can try specifying the location of the private key file with the -i flag, e.g. ssh -i ~/.ssh/id_rsa foobar@123.456.789.
This server is now ready for you to develop and deploy a Node.js web app.
Install Nginx as reverse proxy server
In order to have your web app appear when a web browser visits your server on port 80 - the default port for web requests - you will need to set up nginx to act as a reverse proxy “front server”. This will direct all HTTP requests at port 80 to your node.js server on port 8080.
Other advantages to using nginx are:
- easy serving of cached static files, rather than running dynamic scripts in response to each request
- easy setup of https instead of regular http for secure/encryptic requests/responses
- easy set up of load balancing, where multiple instances of your server share the burden of responding to requests
The following commands install the nginx web server software on CentOS and set it up as a reverse proxy.
Install the EPEL repository, which incluldes “extra packages for enterprise linux”, including nginx:
sudo yum install -y epel-release
Install nginx:
sudo yum install -y nginx
Start nginx by using systemctl a process manager for system-related tools:
systemctl start nginx
Verify nginx starteed:
systemctl status nginx
Have nginx start automatically whenver the system reboots:
sudo systemctl enable nginx
Determine your server’s IP address using ip addr or ifconfig - from either command, you will see several IP addresses output by. Ignore the loopback address, which simply represents the address of your server when it talks to itself, and look for the ethernet eth0 addresss:
ip addr
If you now use a web browser to navigate to http://your_ip_address, you should see a message that nginx has been installed and is running.
By default, nginx serves files located in the /usr/share/nginx/html directory. You can place files in that directory to publish them. To change this server root directory, edit the configuration in /etc/nginx/conf.d/default.conf.
Nginx can host multiple web servers, known as Virtual Hosts. To set up virtual hosts, add new configuration files to /etc/nginx/conf.d - any files in this directory named *.conf are loaded when nginx is started.
References
These notes are based on a few sources: