As of writing, it is 2019 and the buzzwords in the last year or two have been ‘JAMStack’ and 'serverless'. I was transfixed by this concept when I first stumbled upon it, and the possibilities it offers are worthy of consideration. A big part of the JAMStack is the concept of everything being serverless which appeals to me on many levels. The concept is deceptive in that there are still servers, but they are managed by other, more capable people, and I don’t have to get my hands dirty in the world of server management. It can, for example, make deployment of a static website so fast and trivial, that it blows away one of the hurdles that stand between some people and their digital aspirations.
This is all well and good, but whilst the possibilities are continually expanding, depending on what your project is, you may need to spin up a server, make sure it’s secure, host your website or app on it. Incidentally I'm calling it a server which is synonymous with web host, but it's just a Linux machine, which has many other uses. The big advantage is that it is always on, so you can schedule jobs to run in it all day everyday, for data processing, or a development server, for example.
I had never done this before, but I decided to give it a go. It wasn’t nearly as scary as I thought it might be, and was actually pretty fun and I’d recommend that if you are technically minded in any way, you should try setting one up. It may show what is possible with a relatively uncomplicated set up.anyone do it at least once.
In case I need to do it again, here is what I did.
In order to do this it would be helpful if you were familiar or comfortable with the command line, as that is where most of this tutorial takes place. Also, I am doing all this from an Apple MacBook. If you are on Windows, you might need a different tutorial.
First things first, you need a server. A virtual server to be precise, and after hearing good things about Linode from a number of different sources, I set up an account with them, and chose their smallest server (called a Nanode).
To set up a server, you need to choose:
- a distribution - this is the flavour of operating system you desire. If you have a particular flavour you love then choose that but as the foundation for this tutorial, you should know that I chose Ubuntu 18.04
- Server location - If this is to server customers, then choose a location near to them. I am European, and it’s for me, so I chose accordingly.
- Linode Plan - This is the size of server you want. I wanted to mess around on a development server, so I chose Linodes smallest size, a Nanode, which costs $5 per month and gives me 1 CPU, 25Gb of storage and 1 Gb of memory, which is plenty. If I outgrow this, I can easily resize my server. You will also need to enter a root password which you’ll need later on to get in to the server.
So I’ve chosen my server, and after a short while it’s up and running. Now I need to get in to it. To do this I am going to SSH in to it. SSH stands for secure shell, and it provides a way to access a server from the command line of your local machine.
Your new server will have an IP address which can be found in the Networking tab of your Linode dashboard, e.g. 111.22.333.444. If you go to your command line and type ssh firstname.lastname@example.org you will be asked to enter your root password you entered earlier. Once that’s done, you’re in! You should notice that your command line prompt will change. Now everything you do in that terminal window will be giving commands to the server, not your machine. For example, if you type ls to list the directory, it will show the directory of your server, not your local machine. You get the idea. To return to your local machine, type exit.
First things first, make sure your server is up-to-date, by typing in the following:
apt-get update && apt-get upgrade
Next, update the host name and set it to be the name you want to give to your server.
hostnamectl set-hostname serverless-server-1
You can check that it has updated by typing
And then the host file can be updated to associate the updated host name with the IP address of the server. Open the hosts file in a text editor, e.g. nano, by typing:
and add a line with IP address and host name
Setting up a limited user
At the moment, we’re in the server as the root user, which is currently the only user able to login to the server. On top of this, the root user is the most powerful, having unlimited access to the server. Every server has a root user, and depending on how strong your password is, it may not be that hard to access the server if someone really wanted to.
To make things a bit more secure, the first thing to do is add a new user, with limited access rights.
… and add it to the sudo group to allow superuser access
adduser prettystatic sudo
Easy! Now logout of root and login as the new limited user.
Next, instead of using a password to login, we can set up ssh key pairs that are exponentially more difficult to crack than a password.
SSH Key Setup
Add ssh folder to server
mkdir -p ~/.ssh
On local computer generate keys
ssh-keygen -b 4096
… and copy the public one to the server
scp ~/.ssh/id_rsa.pub email@example.com:~/.ssh/authorized_keys
Change some permissions on the server for the SSH
sudo chmod 700 ~/.ssh/ sudo chmod 600 ~/.ssh/*
Great. New user, with SSH key pair authentication.
Finally, the server configuration can be adjusted to only allow access via ssh key pairs, disable password authentication and to cover all bases, root login can be disabled.
sudo nano /etc/ssh/sshd_config
PasswordAuthentication and set it to
Now restart ssh for changes to take effect
sudo systemctl restart sshd
This server is much more secure now that we can be almost certain that no-one else can access it.
This server is going to be used to host and serve websites, and a firewall can be set up to dictate exactly what incoming and outgoing traffic is allowed. On Linux, there is a configuration tool called ufw which stands for uncomplicated firewall. Let’s install it.
sudo apt-get install ufw
And set up some basic rules
- Allow outgoing traffic
sudo ufw default allow outgoing
- Deny incoming traffic
sudo ufw default deny incoming
- Allow ssh
sudo ufw allow ssh
I’m going to be using this as a Jekyll development server which has a default port of 4000 that I want to allow traffic to and from
sudo ufw allow 4000
And then http and https incoming traffic will be allowed
sudo ufw allow http sudo ufw allow https
To enable all those changes…
sudo ufw enable
… and check the status
sudo ufw status
Setting up nginx turns your linux server into something that will serve. It's worth having it set up and ready to be configured to be a development server. If you need to add a custom domain and SSL certificate and dedicate this server to serving production websites, then Nginx is what you want, but that is not covered in this post. For now, as long as Nginx is set up you can serve a website at the IP address of the server.
So, install nginx
sudo apt install nginx
Check what configurations are available
sudo ufw app list
You should have
Nginx Full Nginx HTTP Nginx HTTPS
Set the firewall to accept the HTTP configuration seeing as this is a dev server
sudo ufw allow 'Nginx HTTP'
Nginx sets the default server block to be served from
This is not a comprehensive guide on how to manage and optimise servers, but it is hopefully enough to show what someone might need to do to get themselves a small, cheap server, set it up with some piece of mind around its security and use it for whatever purpose they need.