Nodel

Title: Node.
js API
Edition: I – February 10, 2020
Genre: Software, Web Development
Publisher: Learning Curve Books
Imprint: Independently published
ISBN: 9798612518297
Author: Greg Sidelnikov ([email protected])
Editors, volunteers, contributors: Grace Neufeld.
Primary purpose of Learning Curve Books publishing company is

to provide effective education for web designers, software engineers and all readers
who are interested in being edified in the area of web development.
This edition of Node.js – Server Setup was created to speed up the learning
process of setting up the JavaScript-based Node server for building your own
applications.
For questions and comments about the book you may contact the author or send
an email directly to our office at the email address mentioned below.
Special Offers & Discounts Available
Schools, libraries and educational organizations may qualify for special prices.
Get in touch with our distribution department at [email protected]
Learning Curve Books is trademark of Learning Curve Books, LLC – an in-

dependent book publisher. License is required to distribute this volume in any
form regardless of format. All graphics and content is copyright of Learning Curve
Books, LLC. unless where otherwise stated.
©2018 – 2020 Learning Curve Books, LLC.
0.1. NODE.JS API – WHAT IS THIS BOOK ABOUT?
0.1 Node.js API – What Is This Book About?

This book is an introduction to building your own Node server from scratch.
Although Express is a popular framework that acts as a wrapper around native
Node architecture it will not be covered by this book. There are already plenty of
good tutorials and books that cover the subject.
Node API – as the title suggests – will focus on the bare bone server architecture
written directly on top of Node in vanilla JavaScript.
This book is for those who enjoy the process of understanding how Node-based web
applications work from the ground up, without adding more layers of abstraction.
There are plenty of lessons to be learned here. Even for those who choose to later
move on to Express, MERN or MEAN stack architecture.
Writing server-side applications cannot be done with Node alone. At one point
you will need to talk to a database to store user data, emails and passwords, etc.
For no particular reason a decision was made to use MySQL module. If you are
already familiar with MongoDB it shouldn’t be difficult to rewrite API functions
to support it. What this book will focus on is overall API server architecture.
Even though many diverse subjects will be covered, I think it’s safe to say that
ultimately this is a book for beginners. Someone who has a first time encounter
with server-side programming in JavaScript.
Node API is a book that can be read as a tutorial walk-through: short chapters
contain setup or code required to move on to the next step. So it’s important that
the book is read from start to finish in linear progression.
Node.js
0.1 Node.js API – What Is This Book About? . . . . . . . . . . . .
1 bash.exe 1
1.0.1 What is bash.exe? . . . . . . . . . . . . . . . . . . . . . 2
1.0.2 Ubuntu on Windows? . . . . . . . . . . . . . . . . . . . 3
1.1 Migrating From Apache to Node . . . . . . . . . . . . . . . . . 4
1.2 Node Libraries and Frameworks . . . . . . . . . . . . . . . . . . 4
1.3 Running Node Server . . . . . . . . . . . . . . . . . . . . . . . 5
2 Installation 7
2.1 Mac . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.1.1 apt install nodejs . . . . . . . . . . . . . . . . . . . . . . 8
2.2 Windows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.3 Installing Node on Windows . . . . . . . . . . . . . . . . . . . . 9
2.4 Configuring Node Server Globally . . . . . . . . . . . . . . . . . 11
2.5 Enable Node To Run From Any Directory . . . . . . . . . . . . . 13
2.6 Exit from Node . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.7 Congratulations! . . . . . . . . . . . . . . . . . . . . . . . . . . 15
3 Running Applications 17
NODE.JS
3.1 Project Home Directory . . . . . . . . . . . . . . . . . . . . . . 18

3.2 Running a file via Node . . . . . . . . . . . . . . . . . . . . . . 18
3.3 The process Object . . . . . . . . . . . . . . . . . . . . . . . . 20
3.4 Adding Packages To index.js . . . . . . . . . . . . . . . . . . . . 21
3.5 Next Steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.6 Running Application Server Continuously . . . . . . . . . . . . . 23
3.7 Congratulations! . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.8 Serving Files Continuously . . . . . . . . . . . . . . . . . . . . . 26
4 Building The API 29

4.1 Building Endpoint API . . . . . . . . . . . . . . . . . . . . . . . 30
4.1.1 Integrating API class into our server . . . . . . . . . . . . 31
4.2 Install NPM MySQL Module . . . . . . . . . . . . . . . . . . . . 32
4.2.1 Installing MySQL Server . . . . . . . . . . . . . . . . . . 32
5 Setting Up MySQL Server 33

5.1 MySQL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
5.1.1 Install MySQL Server Locally on Windows . . . . . . . . 33
5.1.2 Install MySQL Server on Ubuntu . . . . . . . . . . . . . 34
5.1.3 Show Existing Databases . . . . . . . . . . . . . . . . . 37
5.1.4 Creating A New Database . . . . . . . . . . . . . . . . . 38
5.1.5 Creating A New MySQL User . . . . . . . . . . . . . . . 39
5.1.6 Make MySQL Open For Remote Access . . . . . . . . . . 40
5.2 Table Specimen . . . . . . . . . . . . . . . . . . . . . . . . . . 44
5.2.1 Bird’s Eye View . . . . . . . . . . . . . . . . . . . . . . 46
6 Adding MySQL to api.js 47

NODE.JS
6.1 Asynchronous Back-End Problem . . . . . . . . . . . . . . . . . 49

6.2 Nesting Queries . . . . . . . . . . . . . . . . . . . . . . . . . . 50
6.3 Handling Many Connections . . . . . . . . . . . . . . . . . . . . 50
6.4 Executing A MySQL Query . . . . . . . . . . . . . . . . . . . . 53
6.4.1 Promises . . . . . . . . . . . . . . . . . . . . . . . . . . 53
6.4.2 The Round-trip Pattern . . . . . . . . . . . . . . . . . . 54
6.4.3 MySQL Promise Function . . . . . . . . . . . . . . . . . 55
6.5 Endpoint API Architecture . . . . . . . . . . . . . . . . . . . . . 56
6.6 Helper Functions . . . . . . . . . . . . . . . . . . . . . . . . . . 57
6.7 Sending POST Data To API Endpoint . . . . . . . . . . . . . . 59
6.8 Streaming POST Data . . . . . . . . . . . . . . . . . . . . . . . 60
6.9 A Pattern For Intercepting API End-Points . . . . . . . . . . . . 62
6.9.1 Other Endpoints . . . . . . . . . . . . . . . . . . . . . . 63
6.10 Authenticating API Actions . . . . . . . . . . . . . . . . . . . . 64
6.11 SSL Connections via HTTPS Protocol . . . . . . . . . . . . . . 65
6.12 Let’s Encrypt . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
6.12.1 Installing certbot . . . . . . . . . . . . . . . . . . . . . 69
6.12.2 Installing git . . . . . . . . . . . . . . . . . . . . . . . 70
6.12.3 Cloning letsencrypt . . . . . . . . . . . . . . . . . . . 70
6.12.4 Create SSL Certificate . . . . . . . . . . . . . . . . . . . 72
6.12.5 ACME Challenge . . . . . . . . . . . . . . . . . . . . . . 75
6.12.6 Adding PEM Files To Run HTTPS Server . . . . . . . . 77
6.13 Complete Source Code . . . . . . . . . . . . . . . . . . . . . . . 82
6.14 Building Scalable Applications . . . . . . . . . . . . . . . . . . . 84
6.14.1 Moving To RAM . . . . . . . . . . . . . . . . . . . . . . 84
6.15 Handle Server Clean Up . . . . . . . . . . . . . . . . . . . . . . 86
NODE.JS
6.16 Master and Stage Servers . . . . . . . . . . . . . . . . . . . . . 86

6.17 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
6.18 JavaScript Grammar . . . . . . . . . . . . . . . . . . . . . . . . 88
NODE.JS
What Is A Server?
Just like a server at a restaurant serving a meal, a computer server is something
that serves files over the network via a dedicated IP address.
For many people server setup and configuration are some of the least exciting parts
about being a software developer.
To those who have propensity toward back-end development, it might be the holy
grail of how the application actually works.
To front-end developers, it’s a black box that generates console errors.
If you want to work for a start-up company as a back-end engineer, having deep
knowledge of how servers work is required experience. But even if you are a front-
end developer, knowing how to set up a server will expand your skill set and might
increase chances of getting a Full Stack developer job.
Today it is not uncommon for front-end developers to know how to work with
servers. You’re only gaining if you learn how to set up your own server.
I also think that servers can be exciting for those with entrepreneurial mindset –
it gives you the power to build your own custom applications. Nothing is more
thrilling than having ability to launch the child of your imagination into the world.
Node Server
Setting up a Node server can be broken into three parts: 1] installation, 2]
adding modules and other dependencies using a manifest file, and 3] writing the
minimum required code in main index.js file to launch and run the server.
We will also write our own api.js module responsible for registering users and
authenticating HTTP requests – in order to read from and write to a database.
This enables you to make virtually any application you can think of.
This book contains examples you can simply copy and paste into Terminal on Mac
or cmd.exe on Windows (or bash.exe on Windows 10,) but your are encouraged
to type them manually. It’s the only way to get better at the command line –
the skill that makes you a much more productive software developer.
You might need to get a web host to deploy your server online but the process is the
same on localhost as it is elsewhere. It’s just the hosted solution will additionally
route your server’s localhost environment to your browser via IP address.
Chapter 1
bash.exe
Did you know that starting with Windows 10 you can work in Linux-like bash
environment directly from your Windows computer?
It can be enabled in Windows 10 Control Panel:
Figure 1.1: Check Windows Subsystem for Linux (off by default.)
1
2 CHAPTER 1. BASH.EXE
There is now bash.exe that looks as follows when executed:
Figure 1.2: Running Linux ls (list directory contents) command on Windows.
You can run Linux-based commands on a Windows file system this way.
In this book we will explore installing Node on Mac, Windows and Linux.
It wouldn’t be fair to show console examples for one operating system only. The
chapters will interchange between Windows bash.exe and Terminal on the Mac.
Commands like npm install package name (to install a node package with Node
Package Manager) or node index.js (to run index.js on Node server) are the same
on all operating systems.
1.0.1 What is bash.exe?

bash.exe is based on Project Astoria that was developed to run Android pro-
grams on Windows.
On Linux we have Wine to run Windows commands.
But bash.exe is like Wine in reverse. It allows you to run Linux-based commands
on Windows. Hence, bash.exe is not really the Linux kernel. It’s just a way to
execute the same Linux binaries but on Windows.
3
1.0.2 Ubuntu on Windows?

But wait. Simply enabling bash.exe in Windows 10 settings isn’t enough. This
way you only get basic commands like ls and ssh. And if logging in to your remote
Linux host is all you need you can skip this section.
However, with current installation of bash.exe you won’t yet have sudo, apt,
apt-get and many more. In order to get them, you have to get Ubuntu from
Windows store.
Figure 1.3: Ubuntu for Windows.
Ubuntu for Windows is free. Simply find it in Windows store and download
it. Follow simple instructions to finish installation. Congrats! You can now use
Ubuntu commands on your Windows 10 bash.exe. Can you use another Linux
distribution such as Debian? Sure! Just look for it in Microsoft store. You can
even set it up so that you can run multiple distribution on the same Windows
computer.
1.1 Migrating From Apache to Node

Many people are still working with PHP servers. There are good reasons for this,
and there is absolutely nothing wrong with PHP.
It’s just Node servers are more often used by applications that run libraries like
React, which can speeds development of the UI components by a large %. I think
Node is better even for solo developers building personal projects because you get
JavaScript on front-end and back-end simultaneously.
If you’re coming to Node from PHP you may be familiar with jQuery.ajax and
fetch API calls to scripts written in PHP from your JavaScript front-end. This
means that just to perform a simple action, you had to know two languages!
Having JavaScript on both front and back-end is magical – you will use the same
language to write your server-side and front-end code!
Although I personally still enjoy writing PHP code, in many cases Node can be
faster than the Apache setup, but I won’t go into that discussion.
I do hope that once you set up a Node server and build your first application you
will never want to go back.
1.2 Node Libraries and Frameworks

It’s tempting to just install a library or a framework with one command in Terminal
or cmd (on Windows) and watch your application magically appear in your browser
at localhost or a dedicated IP address.
But by doing this we are neither learning nor gaining experience with overall ar-
chitecture of a server and the logic behind decisions we need to make in order to
put all of the pieces together. If you choose that route, when errors creep up, you
may not understand what hit you and spend hours in debugging.
In this book, we will walk through the process step by step, without the help of
unnecessary modules, libraries of frameworks.
If we can narrow down on the key issues and avoid common pitfalls, I think the
process of setting up a server can be fun and exciting!
1.3. RUNNING NODE SERVER 5
1.3 Running Node Server

There is really one way of running a node server, regardless of whether you are
doing it on localhost or on a hosted web server with your own static IP address.
If you learn how to set it up on localhost, you will know how to set it up elsewhere.
The process is identical. You will only need to let your server know whether it’s
running locally or in production environment. We’ll cover that later in this book.
Some differences may exist depending on what platform and operating system you
are using. In the next section of the book we will cover both Windows and OSX.
Chapter 2
Installation
There are 3 things you should worry about when installing a Node server:
1. Choosing to install Node globally on your Operating System or making it local
to a project directory on your hard drive.
2. Configuring package.json manifest file – it is automatically created by Node
installation, but it’s nice to know how it actually works, so we’ll break it down.
3. Understanding node modules – the default directory where all Node modules
are downloaded when you install them from command line.
If you can understand these 3 key things about Node server, you’re making a huge
step toward getting better at back-end development with JavaScript. We’ll take
a look at each one as we move forward.
Local vs Production Development Environment
You can think of setting up Node as being synonymous with setting up your
development environment. Once you learn how to run the server on localhost,
it can then be deployed to a remote IP address and and ran on actual domain
name for your application. Do not develop directly on production server. If you’ve
never been employed as a software developer, you will probably want to avoid
developing that habit. In work environment all software is developed locally first
and committed to a decentralized version control system such as GitHub. It goes
without saying that your code should be able to work without bugs on localhost
before it is pushed to the production server.
7
8 CHAPTER 2. INSTALLATION
2.1 Mac
You might want to start learning how to install programs from the command line.
Mac & Linux may have different program installation managers. On OSX you
generally use brew, yum or port (See macports.org) command. On Linux it’s
either apt-get or more commonly just apt. Your should probably use apt.
The reason for so many ways to install packages on Linux-based OS’es is due to
many open-source distributions (Debian, Ubuntu, Fedora, Arch, CentOS...etc.,)
each in own unique flavor and version history.
Technically, you can also use apt on Mac OSX, if it’s available. (It should be.)
If you already have apt but you want brew – which some find to be better – on
your OSX, run apt install linuxbrew-wrapper
Now you can use brew to install packages such as nodejs, for example:
Just run brew install nodejs to install Node in Terminal on OSX.
To do the same on Linux use apt-get install nodejs in bash.
Or simply apt install nodejs in bash. There are minor differences between apt
and apt-get, you can use either one. But apt is considered better.
You can also try apt install nodejs in bash. And this is my preferred way.
2.1.1 apt install nodejs

On Linux-based systems these commands install a package or program locally on
your computer, much like running a GUI-based installer on Windows.
All that happens is locating, downloading and copying program files to a directory
on your hard drive.
After installation of packages such as nodejs and mysql you usually have to
configure and run them. We’ll go over the process in this book!
2.2. WINDOWS 9
2.2 Windows
To get started on Windows, hold Window-key and press R.

Type cmd and hit Enter in the box that appears.
The command line window will open, and it looks similar to this:
To see if node is already installed, type node and hit Enter:
In this case I don’t have node installed, so let’s get started!
2.3 Installing Node on Windows
Some older laptops or PCs might still be 32 bit. So, if you’re not sure, run the
following command in the cmd window: wmic os get osarchitecture
Figure 2.1: Running: wmic os get osarchitecture
Looks like my computer is 64-bit, just as I thought!

To install Node on Windows, first head over to the official Node website:
Figure 2.2: URL: https://nodejs.org/en/download/
Just choose Window Installer and 32-bit or 64-bit version. I chose 64-bit one,
because my computer is 64-bit. Yours is too, most likely.
Figure 2.3: Click on the downloaded file (I’m using Chrome browser here.)
This will begin the Windows installation process. Just click Next everywhere.
2.4. CONFIGURING NODE SERVER GLOBALLY 11
Figure 2.4: I chose default path C:\Program Files\nodejs
2.4 Configuring Node Server Globally

Sometimes you want to be able to call node command from any location on your
hard drive. That’s because your root application folder can be located anywhere.
So now that we installed Node with the Windows installer, we should be able to
run node command from command line, right?
Wrong.
We still can’t launch node globally.
In order to run the node command, you have to be in the directory in which Node
was installed, because that’s where node.exe executable file is.
Earlier we installed node into C:\Program Files\nodejs
Navigate to that directory by typing: cd C:\Program Files\nodejs
Now that we’re there, execute node command:
Your cursor will turn to right bracket >

This means node server is running and awaiting your JavaScript code!
Type console.log(”Hello there.”), to see that Node server works (almost) just like
your regular JavaScript console in Chrome. That’s because Node server runs on the
same JavaScript engine (V8) used in Chrome (with just a few minor distinctions.)
This means that we can launch node and execute JavaScript code. You can
feed node an index.js file to execute, instead of typing the code live in the cmd
program. We’ll take a look at that in just a moment.
2.5. ENABLE NODE TO RUN FROM ANY DIRECTORY 13
2.5 Enable Node To Run From Any Directory

In previous section we ran some JavaScript code on an instance of a node server.
But if you create new projects, it’s likely they will be located in different folders
like C:\develop\my app and as we had just seen we cannot run node outside of
its installation folder by default.
The solution is to include path to node.exe to your Environment Variables.
Figure 2.5: Running: rundll32 sysdm.cpl,EditEnvironmentVariables
Running command rundll32 sysdm.cpl,EditEnvironmentVariables will crack

open Windows system variable editor:
Figure 2.6: Upper half of Environment Variables window.

In the upper box of Environment Variables window (where it says user variables
for YourUserName) click on Path and click Edit... button.
In next window, click New and type C:\Program Files\nodejs on a new line:
Figure 2.7: Upper half of Environment Variables window.
Click OK, then click OK again on underlying window.

That’s it! Now we can run node from anywhere on the hard drive:
Figure 2.8: Now you can run node from anywhere.
Type node then type console.log(”Node Anywhere.”) and watch output.

The function console.log returns undefined because it is not an expression.
If you want to learn more about JavaScript, check out JavaScript Grammar
(www.javascriptgrammar.com) for a free copy.
2.6. EXIT FROM NODE 15
2.6 Exit from Node

After entering some JavaScript, we’re still stuck in Node.
To exit, hold Ctrl and press C key two times in a row.
2.7 Congratulations!
You have just installed Node on your computer.
Chapter 3
Running Applications
So far we only installed a dormant node.exe on the hard drive.

But how do we actually run our own application that serves index.html or any
other file on the system when the browser makes a request to it?
We’ll get to that in just a moment. But first, let’s set up our project directory.
It can be located anywhere on your computer – choose wisely.
17
18 CHAPTER 3. RUNNING APPLICATIONS
3.1 Project Home Directory

This is my favorite part. We’re going to choose where our Node application is
going to live on the hard drive.
I know we installed Node globally on the Operating System, but the project will
usually be located in a local folder, like C:\app
Let’s navigate to C:\ drive using cd C:\ command:
Then, use mkdir app to create new directory C:\app and enter it with cd app.
3.2 Running a file via Node

Instead of typing all the code on the command line, we can pass a JavaScript file
for node to execute. Let’s create our first file containing some basic JavaScript.
Create new file index.js as in C:\app\index.js and type following in it:
You don’t have to import or require any modules to use process object.
It is available in the module scope of your application by default.
3.2. RUNNING A FILE VIA NODE 19
Let’s run our index.js file via node to see what happens. Make sure you are in
C:\app and type the following command in cmd:
Figure 3.1: Run: node index.js
node index.js must be executed from same folder where you have index.js file.
The only reason we can run node (actually node.exe) from this folder is because
we added its installation path to our environment variables in an earlier step. So
if you get an error, make sure to complete that step first!
This runs the single line of code console.log(process) from our index.js script.
See the output on the next page.
3.3 The process Object

In our index.js console.log(process) outputs the contents of process object:
Figure 3.2: Printing out contents of process object.
This Node process object contains all kinds of useful configuration properties.
For example process.versions.v8 tells you V8 engine version it’s running on.
3.4. ADDING PACKAGES TO INDEX.JS 21
3.4 Adding Packages To index.js

Node comes with about 349 packages by default. They are also called modules.
Modules are stored in your node installation folder (C:\Program Files\nodejs
in our case) under node modules.
You might also find them in nodejs\npm folder:
C:\Program Files\nodejs\node modules\npm\node modules\
In browser-side JavaScript code, you can use import / export keywords. But in
addition to that, in node you can use require keyword to add useful modules to
your app from the main NPM repository.
Node assumes path module exists in node modules directory. If not, you can
install it with npm (node package manager ) by running: npm install path
The package called path gives us ability to work with URL paths in an easier way.
Once included path.extname method can parse extension part of a filename from
a string. This is a frequent need in server-side programming:
Update index.js file with above code and run node index.js command:
Figure 3.3: Run: node index.js
Figure 3.4: Output: .js

3.5 Next Steps

We’ve just experimented with running JavaScript files with node, imported a pack-
age, and ran a method. You can run any valid JavaScript file this way.
But there is still a problem. After executing index.js the node process quits, and
control is given back to cmd. This won’t help us run our application server. We
need to build something that runs continuously so we can serve a file every time
someone will access our server from their browser.
3.6. RUNNING APPLICATION SERVER CONTINUOUSLY 23
3.6 Running Application Server Continuously
To continuously listen for connections we need to write server code in index.js.

But don’t worry, the bare bones version is not as difficult as it may sound.
Adding let http = require(”http”) to the very top of index.js is the first step.
When you assign a variable to a package name this way, we make http variable
point to the default object defined in that package. Most packages export a single
object this way, with all the properties and methods already attached to it, so we
can start using them, without worrying how they were implemented!
Likewise, the http object can be used to start listening for requests.
Calling http.createServer makes our node program enter listening state. How
this actually works internally is tied to how your Operating System implements
socket connections. But to get started quick we don’t need to think about that.
Figure 3.5: The bare-bones Node server.
Copy code above into index.js and let’s run it via node index.js command:
Note that we used 127.0.0.1 (It is the same as localhost) with port 3000.
3.7 Congratulations!
There is now a back-end server running from your computer.

Keep cmd window open and paste http://127.0.0.1:3000 into your browser’s
Address Bar. I just did that 3 times and here is the server receiving the requests:
The request shows up as / because we’re accessing just the IP address. It maps
to the root directory of our application.
Now enter http://127.0.0.1:3000/index.html and watch cmd window:
The file index.html doesn’t even exist yet. All we do is intercept incoming requests
and print them out.
Because we are not sending anything at all back to the browser, you will be stuck
on this page ”waiting for 127.0.0.1...”:
3.7. CONGRATULATIONS! 25
Now you know what happens when you try to go to a website and that little note
in the lower left corner appears!
The bar is spinning and it’s only a matter of time until you get the following:
Request timed out. But you can terminate the server yourself with Ctrl + C:
Every time you want to terminate the server manually use this combo. This can
become useful if your server is stuck in an infinite loop or processing a long script.
3.8 Serving Files Continuously
We’ve just set up our Node server to listen to URL requests.

To actually serve the requested file, we need to write additional code.
3.8. SERVING FILES CONTINUOUSLY 27
Update index.js with code above and request 127.0.0.1:3000/index.html or

localhost:3000/index.html from your browser again.
If index.html exists in the root directory of your project (C:\app) then it will be
served to your browser and its contents will be displayed.
But if the requested index.html doesn’t exist the browser will serve an empty
page by inserting blank <html></html> tags on its own for a clean slate view.
Adding More MIME Types
We now have our Node serving files of .html type. If any other type is requested,
the file will be actually downloaded into the browser, instead of shown in the
display view.
This is because if the file extension is not on our mime list, we serve it as
application/octet-stream which means a binary file download.
In order to add other files (I assume you want at least support for .css, .png, .jpg,
.txt, .js and .json files to start doing anything meaningful with your server) all
you have to do is add them to mime variable:
Conclusion
From here on you will build out the rest of your server architecture. The remainder
of this book will walk you through how to install MySQL and write code for
implementing your own API. You can then use this API to build your application.
Chapter 4
Building The API
So far we learned how to serve requested files. But if you want to build real web
applications, you might want to integrate support for API endpoints.
An endpoint is a URL that follows a special pattern. We can intercept this pattern
and upon its presence we will execute an API command, instead of serving a file.
This API command is likely to establish a database connection and either modify
or fetch some type of data. Our server can return a JSON object containing
requested data.
We will take a look at how to create some API end points for our existing Node
server. And then we will install MySQL and execute some queries. Using this
pattern you can build out the rest of your API.
Here is an example of an API endpoint that gets user info:
Here is an example of an API endpoint that sets user info:
When this type of URL is requested, we will cancel serving the file, run some
server-side commands and return a result.
29
30 CHAPTER 4. BUILDING THE API
4.1 Building Endpoint API

Let’s write our API code.
Place the following code into a separate file api.js. It will be imported into our
main server application from index.js.
Static function API.catchAPIrequest will return true if the request.url matches

our API pattern. We will cancel serving the file as usual in this case.
The static API.parts property stores the parts of the API call in an array. For ex-
ample localhost:3000/api/user/get becomes: [”api”, ”user”, ”get”]. Based
on these values we can tell our server what to do (execute MySQL query, etc.)
Static function API.exec will execute the command.
4.1. BUILDING ENDPOINT API 31
4.1.1 Integrating API class into our server

Now that we have the API class ready, let’s add API functionality to index.js.
The changes are minor, they are highlighted below.
Add the following line to the top of index.js on line 004:
This makes our API class available for use in index.js.

Now go to line 026-027’ish in our current index.js file and add lines in green:
We used boolean result of API.catchAPIrequest to branch out.

Our server can now draw a distinction between an API call and a file request.
32 CHAPTER 4. BUILDING THE API
4.2 Install NPM MySQL Module

Before we move on, make sure to install npm module mysql by executing the
following command in your cmd or Windows Terminal or bash.exe on Windows
(Yes, Windows 10+ has its own Linux-like bash.exe) or Terminal (OSX/Linux):
Figure 4.1: Install MySQL by executing npm install mysql on cmd.
I already have mysql installed, so your output might be a bit different.

You might have to type Y and hit Enter a couple times during mysql installation.
This gives us ability to include mysql as a module in our server-side .js scripts:
4.2.1 Installing MySQL Server

In addition to this, you must have MySQL server running on your system or at
some remote location (such as your web host’s IP address: XXX.XXX.XX.XX). If
you’ve already done that step, you can skip the next section.
If not, the next section will demonstrate how to set up your own database running
either on your localhost or at some domain name.
Chapter 5
Setting Up MySQL Server
5.1 MySQL
You can run a MySQL server directly from your computer on localhost – the
home address of your computer as part of your local development environment.
Or you can run it on a remote server (a paid web host, for example.)
5.1.1 Install MySQL Server Locally on Windows

Remember, you have to install MySQL – the actual program to run on your
computer. But you also have to install the accompanying mysql NPM module so
your Node server can include it using require keyword.
If you want to install MySQL on your computer as part of your local develop-
ment environment, one of the fastest ways to do that is to download it from
https://dev.mysql.com/downloads/installer/
Or you can install MySQL together with the Ampps server. You can download it
from https://www.ampps.com/ Just download it, go through the installer and
in under a few minutes you will have MySQL running.
I won’t go into great detail on pressing Next button on installation screen here.
The process is pretty straightforward.
33
34 CHAPTER 5. SETTING UP MYSQL SERVER
5.1.2 Install MySQL Server on Ubuntu

The apt-get (or simply apt) command is perhaps the most popular, because apt
also identifies dependencies, downloads and installs them, whereas dpkg does not.
This section will walk you through the standard MySQL server configuration on a
Ubuntu server, for no particular reason. It’s just a personal favorite.
In the following images, I already used my Mac Terminal app to log into my hosted
web server in the cloud. But if you are running Ubuntu on your computer, the
steps are exactly the same, except you don’t have to log in to your web host.
To log in to your web host open Terminal on your Mac and execute the command
ssh [email protected] and replace the X’es with your actual remote server
address. Enter password, and you’re in!
Once there, you can start executing the following commands.
Make sure all packages we will install are up to date:
Figure 5.1: sudo apt-get update
This will have your Linux system updated with fresh URLs to latest packages. So
when you install mysql in the next step most recent update will be used.
Install the MySQL server on your system:
Figure 5.2: sudo apt-get install mysql-server

5.1. MYSQL 35
Installation will begin and files will start downloading to your hard drive.
At this point you will see a stream of commands flowing on your screen.
Type Y / yes and press Enter if you are ever asked any questions.
Allow incoming and outgoing traffic to/from mysql program:
Assuming mysql installation was successful, you can now use the ufw utility
program (Which stands for Uncomplicated Firewall.)
Figure 5.3: sudo ufw allow mysql
Note, our command starts with sudo which simply runs any program (ufw in this
case) using security privileges of another user.
sudo originally meant superuser do because on older Linux versions it was de-
signed to run only as superuser (usually root). Nowadays, you can use it to
execute commands using any username/password pair on the system. The mean-
ing of sudo acronym was gracefully re-formed substitute user do for said reason.
Uncomplicated Firewall
You can use ufw allow and ufw disallow commands to allow incoming and
outgoing traffic from a program. Without any additional options it is assumed you
are giving full access rights to this program. The ufw command has many other
options, and you can even allow connections on a specific port. Look it up online
if you need the complete documentation.
Figure 5.4: systemctl start mysql (start mysql process)

If you run just the mysql command now, you will enter mysql prompt.
But... it’s important to use mysql -u root -p instead.
This logs you in as root user, which gives you special privileges.
Figure 5.5: mysql -u root -p
You will be prompted for your MySQL root user password. Enter it!
Then you should be greeted with mysql >:
Figure 5.6: mysql -u root -p
Welcome to mysql prompt. This means your MySQL server is running and it’s
ready to receive MySQL commands. Every standard MySQL command you can
execute via code will work in this prompt as well.
5.1. MYSQL 37
5.1.3 Show Existing Databases

When mysql was installed it automatically created information schema – the
first ever database added to your MySQL server. It contains additional information
about all your databases. You can query it, but you cannot change anything on it.
You can type SHOW DATABASES; to see all databases that currently exist on
your MySQL server. On my machine, I already created lavacode and myserver:
Figure 5.7: SHOW DATABASES;

5.1.4 Creating A New Database

To create your own database, run CREATE DATABASE database name;
(Don’t forget to end your statement with ; or the mysql prompt will drop to the
next line without executing the command.)
Figure 5.8: CREATE DATABASE myserver;
You can also check current users by running Select User FROM mysql.user;
Figure 5.9: SELECT user FROM mysql.user;
Note that mysql command keywords and table names are case insensitive. You
could have typed either Select User or SELECT user to the same effect!
5.1. MYSQL 39
5.1.5 Creating A New MySQL User

In addition to root user, we can create other users by issuing following command:
CREATE USER ’felix’@’localhost’ IDENTIFIED BY ’PassWord55723!’;

But now we also need to give this user access privileges using GRANT command:
GRANT ALL PRIVILEGES ON *.* TO ’felix’@’localhost’;

This gives felix admin-level privileges on all tables in the database (*.*)
Now we have user felix ready to log in to the database via localhost address with
provided password. This is the account you can use in your JavaScript code when
we get to the Chapter 6: Adding MySQL to api.js
We just need to make sure our database is open for remote access. See next page!
5.1.6 Make MySQL Open For Remote Access

Enter nano /etc/mysql/my.cnf in your Linux bash to open your default MySQL
configuration file with nano editor.
Usually /etc/mysql/my.cnf is the main configuration file. But it’s possible that
it might not be, depending on Linux and MySQL version, etc:
If your my.cnf config looks like above, you need to find the actual mysql config-
uration file which should reside in one of the folders specified with !includedir.
Snoop around these folders to find the file with a similar .cnf name. In my case
it was mysqld.cnf, so I entered the following command to edit it:
nano /etc/mysql/mysql.conf.d/mysqld.cnf
5.1. MYSQL 41
Note below comments the file starts with [mysqld safe]. This is the recom-
mended way to start MySQL server on Unix. On Windows you will probably have
just [mysqld] in your config file, which may be called either my.ini or my.cfg.
[mysqld safe] tells MySQL server to restart in case of an error, among some other
helpful features. If you see this command you’re in the right MySQL config file.
Hold control, press W, type bind-address and press Enter.

If you are in the right MySQL config file, your cursor will jump to this line:
Comment the line bind-address = 127.0.0.1 by adding # to the front:
Now that the line is commented...

Hold control and press O, then press Enter to save .cnf file:
Type /etc/init.d/mysql start or /etc/init.d/mysql restart in Linux bash.

This will restart the MySQL server using the new configuration file.
Congrats, your server will now accept connections remotely.
This means from now you can log into your MySQL server using a MySQL client
such as Sequel Pro on Mac or MySQL Workbench on Windows, using the login
token pair felix/PassWord557123! we created earlier.
5.1. MYSQL 43
Final Words
Whenever you start a new project, if a hosted database is the only option, you will
probably want to open your MySQL database for remote access. This way you
can access it from your localhost environment without having to install MySQL
server on your machine.
However, this is usually considered bad security practice. At one point, you will
want to limit your MySQL database access to localhost only. This means only
your Node server can connect to the database directly.
This still doesn’t prevent database injection attacks.
Injection attacks use your API endpoints to pass actual MySQL query code into
one of the values sent together with the POST or GET requests.
By clever use of quote characters the query can be re-constructed to execute a
different MySQL command which can be virtually anything.
Keep this in mind when writing your endpoint code and generally thinking about
how to secure your database!
Writing MySQL Code
Now that we have our mysql server running and waiting for incoming user connec-
tions, we need to write some code on our Node back-end (index.js) as a starting
point for our MySQL-based API.
In the next chapter of this book we will add new mysql code by starting a connec-
tion to the server and executing a query. From that point on you can extend the
API by implementing the entire CRUD pattern.
CRUD
We already started api.js file earlier in this book, which – at this point – can only
interpret certain URL patterns as API endpoints and cancel serving the file.
But we are still not making any MySQL connections to the server. Nor are we
executing MySQL queries to actually Create, Read, Update or Delete rows –
collectively known as the CRUD API.
Once you implement the CRUD on your server, you will be ready to build out the
rest of your application. Usually the CRUD gives you ability to code pretty much
any feature on the front-end.
5.2 Table Specimen

In order to execute queries on a MySQL database, we need to create a test table.
For examples that follow we will assume that user table already exists.
You can create it via command line, but it is recommended to get good UI-based
database manager software such as Sequel Pro on Mac or MySQL Workbench
on Windows. As your application-data tables begin to evolve editing them from
command line might turn into Hell.
Because our MySQL server is open for remote connection, it’s possible to log in
to it from Sequel Pro using felix account created earlier. If you set up MySQL
server on localhost use that as the Host. Otherwise use your web host’s database
IP address (which very often is the same IP address for your web hosting server.
It makes senes, because technically MySQL is installed on ”localhost” there too.)
5.2. TABLE SPECIMEN 45
Once logged into your favorite MySQL editor, create the user table that looks
similar to the following design:
Perhaps your table can be completely different. But for users at least have a
username and email address columns!
5.2.1 Bird’s Eye View

Before we move on to the next chapter, I want you to take a look at the bird’s
eye view architecture we’ve got going on so far.
This is the mental model you will eventually arrive at after breaking down a simple
Node server into compartments.
Our barebones server requires only http, path, fs and mysql modules.
Note that the mysql object from MySQL package will return result in a separate
process in a callback function. So it’s a bit different from how you would expect
MySQL to work on a PHP server.
If we can find a way to handle the result and write back to the requesting client from
the callback, we can make API calls just as you would expect from Apache / PHP
setup. Think of it as an extra step. This makes Node MySQL API architecture a
bit more challenging, but also more flexible.
Chapter 6
Adding MySQL to api.js
Now that we have our database running either on localhost or at some remote
location, we’re ready to integrate our MySQL code into our server, so that we can
convert our API endpoint calls to MySQL queries.
Go to our app.js file and add lines highlighted in green to the very top of the file
(so that now we will have database and API classes together in the same file).
If you think class database should be Database, go ahead and change it to your
heart’s desires! The only reason it’s in lowercase is because we’re only using static
methods here, and in my opinion database.create looks more elegant.
47
48 CHAPTER 6. ADDING MYSQL TO API.JS
Make sure to replace XX.XX.XX.XXX with your actual web server IP, or if you’re
running Node server locally localhost (or 127.0.0.1).
Replace other XXXXX entries with your actual database username (felix), pass-
word (PassWord55723!) and database name – all of which were set up in the
previous section – or just use your own values.
The static function create() can now be used to create a database connection.
You can call database.create() once from the very top of index.js after require
6.1. ASYNCHRONOUS BACK-END PROBLEM 49
statements. Your server will connect to the MySQL database and you will be ready
to start executing MySQL queries.
Inside static function exec all we do is call connection.query method on our
database connection object.
6.1 Asynchronous Back-End Problem
If you are coming to Node from Apache PHP server, it is inevitable that you will
experience a paradigm shift, when trying to write your own mysql queries.
In PHP, in order to execute a mysql query, all you had to do was write a script
such as get user.php on the back-end.
Then you would use fetch API or a jQuery.ajax to call that file. The result would
be generated as part of the page content and returned inside the callback. This
way you could safely update the front end UI. We used to call this asynchronous
HTTP requests, or ”ajax” calls. It works naturally for updating the front end!
A Node callback doesn’t run in the same process with the function serving the file
to browser. So the file will be served before the mysql query returns! And this is
the fundamental problem we need to solve in order to build a decent API server.
Executing a MySQL query works exactly the same on Node. The result is returned
via a callback function in a separate process. And that’s part of the problem,
because we’re not on front-end anymore.
A callback will often depend on completion of another MySQL query. So we will
also need a way to execute a chain of multiple queries. And this can introduce a
whole new set of problems.
Solution. If we can somehow forge the process of serving the file with the result
we get from mysql query, we’re taking first step toward solving this problem.
One way of solving this problem is to use promises.
6.2 Nesting Queries

First, let’s take a look at what problem we’re trying to solve with nested queries.
Associative Tables
Let’s say you want to store followers of all existing users. You can’t store users
and followers in the same table because they are two completely different types
of data sets.
However, you can store followers in a separate table and ”associate” it with users
by adding userID column to it.
This way when you want to access an entry in followers table you can use the
userID column to look up which user it belongs to.
Nesting Queries
For the reason stated above, you will often need to write chains of nested queries.
To get followers from a user account, first you will retrieve a user from users
table. Then you will retrieve followers associated with that user’s primary key ID.
Because followers table has a link to the user via the user id column, you should
be able to gather a list of followers separately for that individual user.
6.3 Handling Many Connections

Default server configurations start out with some arbitrary connection limits. For
latest versions of MySQL this limit is 151 maximum connections.
If your app connects many users at the same time, this limit will be reached and
your program will cause the famous ”too many connections” error.
Figure 6.1: ER CON COUNT ERROR

6.3. HANDLING MANY CONNECTIONS 51
This is just a MySQL error caused by configuration limits. It doesn’t actually

mean your server can’t handle that many. In general, these limits are set by
default because every time you initiate a connection, a new thread is created.
Creating a new thread process just to serve 1 user will inefficiently use up your
server’s memory resources. Of course, once we no longer need it, this thread is
eventually terminated by our code. But with thousands of users connecting every
other second you are creating anywhere from hundreds to thousands of threads at
the same time hogging the CPU and possibly the hard drive.
When developing your server you might run into configuration limits on your
MySQL database.
To find out the maximum number of allowed connections, enter the mysql prompt
(by typing mysql either on mac or PC – it’s the same command) and enter the
mysql command: show variables like ”max connections”;
Figure 6.2: show variables like ”max connections”;
Don’t forget the semicolon at the end of each mysql command or it will not run.
I’m running set global max connections = 300; to double connection limit:
Figure 6.3: set global max connections = 300;
And if everything went well, running show variables like command should yield the
following result:
Figure 6.4: show variables like ”max connections”;
We’ve just expanded the theoretical number of simultaneous MySQL connections.

Creating MySQL connections is computationally expensive. Outsourcing the server
workload by increasing the number of connections might work but it is still a naive
strategy. And often, just one connection is enough to handle thousands of queries.
Having said this, before increasing the number of connections, make sure your
system has enough resources to handle them.
However, this doesn’t prevent programmers from using connection pools de-
scribed in next section. Connection pools might offer a fair deal of additional
optimization. Just keep in mind, each architectural decision will depend on the
actual purpose of your API server.
6.4. EXECUTING A MYSQL QUERY 53
6.4 Executing A MySQL Query

So far we explored the idea of executing a MySQL query with exec(command)
method residing on our main database class as a static function.
But architecturally, this won’t do any good as far as our API goes. Remember
that on Node a MySQL query returns in a callback function which produces the
query result value away from our main execution thread.
That’s fine if all you want to do is execute a query or two without caring about
sending a response back to the browser.
But if you want to build an API which deals with a series of unique endpoint
calls, you might seriously consider using promises. They can help you to properly
respond to each incoming request with data returned from a MySQL query.
6.4.1 Promises
Promises add one extra layer to the standard callback pattern. When the Promise
returns, it returns a Promise object, instead of the callback function. This Promise
object is not guaranteed to carry a value yet, until it is resolved at a later time,
which is exactly why it’s called a promise. You can use the built-in .then method
on the Promise object to retrieve the results from the MySQL query.
The Promise’s .then method allows us to wait for the result and resolve it soon
as it becomes available even if it happens at a later time.
Using .then method allows you to avoid the nested callback pattern (or at least
reduce its occurence) where you keep nesting a callback within a callback, because
one API call depends on the data returned from the previous one.
Promise-based MySQL query
So what does a promise-based MySQL query call look like in Node?

Let’s take a look at that in the context of entire API endpoint round trip (Which
includes requesting a resource using an API endpoint and responding to it.)
6.4.2 The Round-trip Pattern
Before we dive into source code, let’s look at how it all fits together, starting from
server listener and finally returning a value from MySQL query.
6.4. EXECUTING A MYSQL QUERY 55
6.4.3 MySQL Promise Function

One of the ways to build the function responsible for executing the MySQL query
related to an endpoint call is by wrapping it in a promise and returning the result
using resolve method - just pass payload to it in JSON format. Here is an
example of just the action register user function, but all other actions (login,
tweet, comment, etc.) follow the same pattern:
6.5 Endpoint API Architecture

Below is a diagram displaying the minimum structure of a functioning Node API
server. It’s the same diagram as you’ve just seen above, except in the context of
a concrete API call to execute user login action.
Logging in should be treated as a special API action. It’s the only endpoint that
performs the initial authentication step.
6.6. HELPER FUNCTIONS 57
This is where you will take user’s password, encrypt it with an algorithm such as
SHA3, and compare it with the SHA3 password already stored in the database
under password sha3 in SHA3-encrypted format.
The primary purpose of logging in is to obtain an AuthToken. This becomes your
key to perform other authenticated endpoint actions – follow, tweet, retweet,
like, comment, update user settings, etc.
6.6 Helper Functions

To avoid rewriting redundant code, I wrote these functions that accomplish various
actions you’ll commonly come across when writing your Node API server. Using
them in various places will help reduce clutter and keep your code clean.
identify(target, action)
target argument is database table name (such as user, tweet, etc.) The action
argument is the ”what” you want to do to target (get, update, delete, etc.)
Figure 6.5: Identify an API endpoint request from API.parts. The API.parts
array is available globally. It represents the API endpoint call broken down into 3
parts. If there is a match, the return value is true.
respond(response, content)
The response argument is the response object from the server and content is a
string representing an object in JSON format (to send back to the browser) which
is usually formed by resolve method inside a promise from MySQL function.
json(chunks)
In Node POST data is streamed instead of received as a single object. So we need

this json function that will convert all received chunks into actual JSON object.
Figure 6.6: Form a JSON object from the streamed POST data chunks.
I intentionally keep my function names short and use common names. This way
when putting it all together the code turns out to be so much cleaner, readable
and easy to maintain down the road.
6.7. SENDING POST DATA TO API ENDPOINT 59
6.7 Sending POST Data To API Endpoint

Usually an API endpoint expects additional data sent together with the request.
For example, if we want to get user data for user with user id of 1, we need
to include that user id in the POST request as additional data specified in JSON
format, for example: {"user id": 1};
When you receive the request object, you only get the header data instantly. The
POST data is submitted separately and you need to listen for it using two different
events: ”data” and ”end” because it’s sent as a stream / buffer.
The request.data event tells you that there is POST data incoming. But instead
of sending entire data object, Node streams it in chunks. So you need to put it
together using chunks first. One way of doing that is to create an array and keep
pushing each chunk received in the ”data” event’s callback onto that array.
The request.end event tells you that all chunks of the POST data have been
finished downloading. At this point you would take the chunks array and convert
it to a JSON object.
This is done using Node’s built-in Buffer.concat( chunks ).toString() method.
The .toString() method is required to convert binary chunk data to string format.
After this you simply convert the resulting string to a JavaScript object using the
native JSON.parse function. We’ll see the source code for that in just a moment!
If the POST data is short, it is possible to receive only one chunk. But you still
need to intercept it as a stream.
6.8 Streaming POST Data

Let’s look at streaming POST data as an isolated example first:
Going from here we can finish building out the entire round-trip pattern.
(See next page.)
6.8. STREAMING POST DATA 61
Knowing everything we do now, let’s upgrade our API class by adding support for
one endpoint to enable User Registration. This endpoint triggers the function we
wrote earlier: action user register
6.9 A Pattern For Intercepting API End-Points

In the code from this book we’ve chosen to let the server know that the URL is
an API end-point by using the word ”api” in the first segment of the URL. Our
back-end is designed to treat requests that follow this pattern as an API action.
All other patterns will attempt to simply download the file. Of course your own
application may not have to do this extra test.
(for example: https://www.twitter.com/user/status/id may look more el-

egant to some. It’s up to your preference to define what this pattern should be!)
6.9. A PATTERN FOR INTERCEPTING API END-POINTS 63
Note that first we wait for POST data stream to finish. Only then we call identify
function to figure out the target and action details for the incoming request.
This request is created by URL that was typed into the address bar (or called
programmatically using the fetch API on the front-end.)
This is our core API end-point logic. Once the code is implemented, all we have
to do is continue adding new end-points to this pattern.
The next few pages will go over source code for some of the most common API
end-points you’ll have to build if you wish to create a user login based application
(which is pretty much all of them.)
6.9.1 Other Endpoints
So far we’ve only covered register user endpoint. It would take too much book
space on printing them all here. The rest of end-points can be found in api.js file
in the GitHub repository for this book:
https://github.com/javascriptteacher/node/
Under: /module/api/api.js
In this file you will find an object called Action with static functions attached to
it as follows:
These methods point to related functions responsible for processing a particular

action. They all return a Promise. The JSON message from each is then sent
back to the back-end to validate that the API action has returned (and whether
it failed or succeeded, which will be indicated by success property.)
You can’t make authenticated connections without understanding the concept of
sessions. Usually a session is stored in a separate database table with a unique
ID, Authentication Token ID, and a link back to the user ID from which the session
was initiated.
So let’s open a database editor (such as Sequel Pro on Mac) and add a new table
called session to existing database we created earlier in the book.
6.10 Authenticating API Actions

The User Login end-point performs a special action of generating a new Auth-
Token, or retrieving one that already exists in a session table.
Many other end-points, such as Update User require an existing authentication
token to be passed together with the request. This makes sense, because without
AuthToken anyone could have made an API call to our server. But we only want
to allow those calls that carry AuthToken in the payload.
This adds bulletproof security to our API, especially if the AuthToken was gener-
ated on a secondary server whose IP address is not known to the end user.
Hence, it is recommended that a completely new server should be running sepa-
rately solely for the purpose of generating the authentication token.
6.11. SSL CONNECTIONS VIA HTTPS PROTOCOL 65
6.11 SSL Connections via HTTPS Protocol

Modern web-applications are not required to be hosted via secure HTTPS con-
nections. However, there is a good reason you should not run production-ready
applications over HTTP. (The way we have been doing it up to this point.)
This means that the connection will be encrypted by something called Secure
Sockets Layer or SSL for short.
SSL helps you encrypt the data sent from your computer to your server. This
helps provide increased security for data such as credit card numbers, emails and
passwords. With HTTP protocol, the data is sent as-is. (Perhaps, it may have
been compressed, but not really encrypted by an encryption algorithm.)
This is important because unless you implement SSL the data sent to the server is
not secure. Also Chrome and other browsers will display ”Not Secure” message
next to your domain name.
So far we’ve worked with Node’s http module. We assigned the variable http to
an import from require(’http’) module on the first line of index.js file.
Luckily for us Node already has a module available to cover https and it’s easy
to switch! Just change ”http” to ”https”:
It’s up to you if you want to change variable name too but we’ll leave it intact so
we don’t break the rest of our code we’ve written so far.
6.12 Let’s Encrypt

Let’s go over setting up free SSL certificates on Linux-based operating systems.
The preferred OS for most web host providers. However, the commands described
in this section are the same in Terminal and bash.exe on Windows.
To use letsencrypt we must update packages, install git (if you haven’t already,)
clone and install letsencrypt repository and run a few commands in Terminal or
bash on Windows 10.
For no reason in particular I’ll use bash.exe on Windows 10 but you can use
Terminal on OSX.
First, I will launch bash.exe from Start menu. The command line window opens:
Let’s login to your web host as root user using ssh command. Just replace
XX.XXX.XX.XXX with the static IP address where you are hosting your website.
You will be asked to enter your root user password:
We should now be logged into the web host. (See server log on next page.)
6.12. LET’S ENCRYPT 67
You will be greeted by a screen similar to this. (I’m running Ubuntu 18.)
To install certbot we need to update the packages first. This is because certbot
developers continue applying improvements and your Ubuntu server may not have
the latest version on initial install.
Run sudo add-apt-repository ppa:certbot/certbot command to add certbot
repository to your Ubuntu server:
Just press Enter and latest packages will be added to mirror files which is a list
of links pointing to the latest version of the packages.
Now run sudo apt-get update to actually download the updated packages:
This step is important - it will update your certbot packages to latest version.
If you find yourself on CentOS or Debian you can do the same thing as follows:
On CentOS run sudo yum update && sudo yum upgrade
On Debian run sudo apt update && sudo apt upgrade
Here the && symbols will cause your packages to be updated followed by being
upgraded without having to execute each command separately.
6.12.1 Installing certbot

Run apt-get install certbot to install certbot package. We’re already logged in
as root, so there is no need to use sudo command (otherwise also prepend sudo.)
Press Y and hit Enter.

Installation log should roll on the screen and after that we should be good to go!
6.12.2 Installing git

In order to install letsencrypt we need to clone the latest version from git. But
in order to do that we first need to make sure we have git installed on our system:
We can now use git command. Keep in mind that in this example we are in
bash.exe on Windows 10. If sudo or apt-get are not working, there is a work-
around.
First, if you are hosting your server remotely, use ssh in bash.exe to log into your
server, and all linux commands will become available over there.
Second, if you are developing on localhost and don’t need to log into your re-
mote host, you must install Ubuntu for Windows in addition to bash.exe (This is
explained in first chapter of this book.)
Third, if you are working directly on Linux, this doesn’t apply and apt-get should
work out of the box. If not, it can be installed manually.
6.12.3 Cloning letsencrypt

Now we’re ready to clone the latest version of letsencrypt to our server. This is
accomplished by running the following Linux command:
git clone https://github.com/letsencrypt/letsencrypt /opt/letsencrypt
Figure 6.7: Again, you might want to precede this command with sudo as shown
in the following example.
Figure 6.8: Again, you might want to precede this command with sudo as shown.
Figure 6.9: Cloning begins...

6.12.4 Create SSL Certificate

Finally, we’re ready to create the SSL certificate that will ultimately enable us to
serve files via https instead of http. Following previous steps type this command
to navigate to newly created letsencrypt directory:
cd /opt/letsencrypt
You will navigate to the folder where letsencrypt was just installed.
(cd stands for Change Directory on Linux-based OS.)
Figure 6.10: Navigating to letsencrypt directory.
Run the following command...

sudo -H ./letsencrypt-auto certonly –standalone -d A.com -d www.A.com
(replace A.com with your domain name)
Figure 6.11: Initiate creation of the SSL certificate for example.com
After this you will see letsencrypt-auto command generating the needed certifi-
cate files with certbot and creating http challenges automatically!
At this stage you might be asked several questions. I’m not going to include them
here to save book space. Just enter your email address when asked and few other
things. Choose (A)gree or (Y)es option every time asked.
On the next page, let’s take a look at what actually happened once the certificate
is generated:
The important parts are highlighted:
Letsencrypt created and verified http challenges (this is needed in order to verify
that the domain name belongs to you, but in this instance this is done for us
automatically.)
Next... it also generated two pem keys for our example.com domain name.
Congratulations! At this point your domain name is https-enabled.
But there is one more thing. Let’s check where the files live and get familiar with
the directory where the keys were actually generated:
Figure 6.12: The certificate files were created in /etc/letsencrypt/live directory.

Let’s output contents of live directory. This is where letsencrypt will store all
certificate keys for all domain names on the server under respective folder name.
Figure 6.13: Run ls command to list contents of the directory.
Inside example.com directory you will find several pem files generated by letsen-
crypt. We only need cert.pem and privkey.pem
We’re ready to start using the certificate. All we need to do is add some new code
to the existing index.js file.
You can skip the next section and jump over to Adding PEM files to run
HTTPS server section.
6.12.5 ACME Challenge
If you completed steps to install the SSL certificate in previous section you don’t
need to do this next step. But in some cases ACME challenges allow you to verify
that you are the owner of domain name.
I repeat, skip it! I just provided it here for reference.
Run certbot certonly –manual or certbot certonly –manual (note this is
actually double dash as shown below, not one dash.)
If you’re hosting multiple domains, you can use the same certificate, so just specify
as many domain names as you need separated by space or comma. In this case
we’re simply adding example site awesomesite.com
Enter your domain name (without www.) and press Enter.
Type Y and press Enter to agree and proceed.

To verify we own the domain name, we need to manually create the file named
afBX9EXvQqUqzFooe02-22EfJ5TzcniFw7CvqlaMXCA – just the part be-
fore the dot (.) highlighted in yellow in screenshot above.
Place this file into your .well-known/acme-challenge directory on your server.
(The root folder of your site.) You might have to create these folders first.
Now edit the contents of that file by pasting the whole string generated by certbot
into it. In our case it is the long text afBX9EXvQqUqzFooe02-22EfJ5TzcniFw
7CvqlaMXCA.rX9ThTxJ4y47aLEr7xgEWcOm4v7Jr5kJeT4PLA98-0
Keep in mind that this filename will be generated every time you run certbot. So
if you’ve already ran this command before it should be changed again.
Press Enter to Continue.
At this point certbot will verify existence of the file. Again, if you used the first
method in this section without acme challenge, you should be good to go.
6.12.6 Adding PEM Files To Run HTTPS Server

Navigate to /etc/letsencrypt/live/site.com/ directory to verify the pem files
were actually generated by steps taken in previous section.
Now that we have privkey.pem and cert.pem generated we need to pass them
to our Node server configuration using an options object.
Let’s add this to our existing server code:
Just replace site.com with your actual domain name.

For https connections it is proper to use port 443.
You might also need to open port 443 on your system for this to start working.
However, this is not a requirement – any allowed port number will still work.
Update the index.js file with code above. Log into your web host. Navigate to
the root directory of your application and run node index. Believe it or not, if all
goes well, at this point your server will be accessible via https protocol, and the
”Not Secure” message in Chrome (and other browsers) will disappear.
Putting It All Together

Now that we are running an https server, we can finish building our API.
In previous chapters we worked out a pattern that can now be used to add more
end-points without modifying main server code.
In a real work environment, you will probably be working with another library (such
as React) or Angular framework. And many software development companies
choose Express as a Node wrapper. However, the principles remain the same.
Knowing how this works at the bare bone level is important if you want to un-
derstand how to architect your own API end-point server regardless of coding
environment. This was one of the primary goals of this book.
This proprietary callback API system would be among some of the very first tasks
assigned to a Full Stack developer working at a new start-up. After all, every
application must have user registration and authentication functionality.
API Code Pattern
To keep our API code organized it’s best that each API end-point follows a familiar
pattern. This way it will be easier to add support for new functionality to our server
by simply adding new function calls.
Whenever you need to add a new custom end-point, simply write a function con-
taining a promise and your MySQL code handling the database reads or writes.
Following exactly the same pattern set in previous chapter.
On the next page we will put together all of the API end-point functions we’ve
created so far into a list.
Here each if-statement will also follow exactly the same pattern. We’ll check the
URI for a specific command and its action and execute related function.
For example user/register, user/login, user/logout and so on...
But we’ll have to make sure that the POST data is available after being streamed
by the server. Let’s take a look at how this can be handled on the next page...
POST data is not available until request produces ”end” event. This is where we
put the chunks together into actual JSON object using the json helper function.
To add support for more endpoints just follow the same pattern. Place them inside
request.on(’end’)’s callback which ensures POST data finished streaming:
The request and response objects come all the way from initial call to
http.createServer(function(request, response)... in index.js. In other words,
in Node they are created automatically by createServer function.
6.13 Complete Source Code

To see the complete source code of the Node server we’ve built so far, you can
fork it from the public GitHub repository for this book:
https://github.com/javascriptteacher/node
Note: Before running this app on your localhost (cmd > node index.js) make
sure you already have a running MySQL server with user table in it and you included
MySQL credentials (IP, username, password and database name) in api.js
Note: This is not a complete application, but it has enough scaffold code to get
you started fast. User API is essential to most web applications. The rest of
endpoints will depend on your app.
Note: You should replace md5 with sha3 if you need to ramp up security for
your user passwords. Use npm install sha3 to install the SHA3 module which
also includes Keccak.
After running node index.js you should see index.html at http://localhost:3000
Press on the API endpoint test buttons and watch console. Remember that each
one of our endpoints returns a JSON object. This object has a property ”success”,
6.13. COMPLETE SOURCE CODE 83
which can be either true or false. The ”message” property will contain detailed
information about what happened.
For example:
{success: false, message: ”user with this ID does not exist.”}
{success: true, message: ”user (felix) was successfully registered!”}
Based on the success value you can update the front-end.
These endpoints are ready for integration into your own front-end UI/UX. You can
call them via onclick attribute or elsewhere from your JavaScript code.
How To Run The Server
I recommend rewriting examples in this book manually from scratch. However,

for a quick demo, you can copy the files from my GitHub repository and run your
very own Node server in under just a few minutes!
Import GitHub files into your project or copy them into one folder.
Navigate to that folder from command line.
Enter: node index.js and your index.html should be accessible via
http://localhost:3000 in your browser.
From here on, it’s up to you how you want to build out the rest of the front-end
and which API end-points you wish to add. This depends solely on the purpose
of your application.
For example if your application requires image uploads, you can create a new end
point and make a function called action image upload.
Multer is a good module that helps with file uploads from an HTML form. To get
it simply run npm install multer command in Terminal or cmd.exe/bash.exe.
6.14 Building Scalable Applications

So far we learned how to build an API Node server that uses MySQL to perform
storage and the classic CRUD data operations.
MySQL, MariaDB, MongoDB and other similar software might work well in the
beginning stages of your application, running on a single server.
You will be able to host under 1K - 10K users (or maybe even more) based on
the amount of RAM available, the IO rate of your SSD drive, and how well your
server code is optimized. Fine tuning your server and writing code optimized for
performance can be magical for increasing the number of users that can be
served by a single instance of your server.
But once your server’s maximum bandwidth is exhausted, it will start coming to
a crawl if any more users join and start using your application.
In the olden days, Battle.net running Diablo I and StarCraft I used to work off
a single server with hundreds of thousands of users connected simultaneously.
This is an example of how limited resources encourage efficient code.
It’s not uncommon for web applications of today to be real-time. Additional
thought and engineering must be applied if you want to build large-scale applica-
tions that serve hundreds of thousands or even millions of simultaneous users.
6.14.1 Moving To RAM
No modern real-time application (Twitter, for example) uses SQL databases to

serve primary features to their users. The reason is quite simple. When using SQL
for everything it means there is a hard drive operation every time someone makes
a request! This can be an overkill, because you are outsourcing the slowest part
of your system to serve key features of your application.
One solution is to switch to an SSD server (and to be honest, most of them already
are by today’s standards.) But what if we store all of our data in RAM instead?
This way we can avoid touching the hard drive completely (by caching frequently
accessed data in an object.)
Of course, the original data will still be stored on the hard drive. The difference is
6.14. BUILDING SCALABLE APPLICATIONS 85
that when you launch your Node server, it will be copied into a JavaScript object.
When client requests a certain resource we can simply serve it directly from that
object. It will contain a copy of the data your MySQL database would contain.
But your application will become significantly more efficient.
You will still write this object to the hard drive, so it can permanently store the
most recent data, but not as frequently as you would with a 100% MySQL setup.
Some databases such as MongoDB also implement similar RAM caching features.
But they still may not be optimized for the primary purpose of your application
which only you know. Thus missing the chance of taking advantage of the full
potential of your server’s horsepower.
And just like with anything else, you can either ”install a package,” or install and
configure existing software, or you can learn to write your own code. Without the
latter, you may never be able to understand or develop the skill of building an
extremely efficient system for serving millions of users.
Unfortunately, this architecture will not be covered in 1st edition of this book due
to time constraints I’ve had within which I had to finish this volume. However, I
plan on second edition where I will go into greater detail on the subject. For now,
the least I could do is to give you a pointer in the right direction.
6.15 Handle Server Clean Up

The server can exit for many reasons. For example, it can time out based on
default timeout value provided in server settings.
There are a few cases in which your server might exit, either spontaneously as a
response to an exception, or when Ctrl + C combo is detected.
You might want to do some code clean up when that happens:
Let’s kill database connection every time server exits for any reason.
6.16 Master and Stage Servers

After a production cycle or coding sprint marathon the new code must be deployed
to the main production server so that your customers can take advantage of the
new features. This is your main public application server.
You should probably set up a secondary stage server where you would deploy the
same code. This is important for two reasons: 1] you can show your application to
6.17. CONCLUSION 87
your client, teammates, investors, friends or others who can test the new version
of the application before the public release, and 2] you can make sure deployment
process doesn’t break the main production server with new unexpected bugs.
The stage server can be used to test integrity of the newly deployed code, without
risking the chance of deploying erroneous code directly to master, losing customers
and having a panic attack.
Now you are thinking like a real production engineer!
Both master and stage are nothing more than two versions of your application
running separately in two different locations. So I won’t talk about how to set up
a stage server – the process is exactly the same. The difference is in where you
push your code to.
6.17 Conclusion
At this point you should be ready to host your own secure SSL applications from
https://www.appname.com if you indeed followed all instructions described in
this book up to this point.
In the next edition I will cover scaling your application and making it accessible by
millions of simultaneous users. There is currently no set date as to when it will
come out, but follow my work at @js tut on Twitter for a future announcement.
I hope that this Node API book was a faithful guide on your journey to becoming
a software developer. Thanks for reading!
6.18 JavaScript Grammar

If you enjoyed Node API, look for other books by Learning Curve Books.
Look for a copy of JavaScript Grammar on Amazon. (cover art may vary.)
Index
adding packages, 21 CREATEUSER;, 39

Apache, 4 CRUD, 43
API, 48, 56, 62, 65 Ctrl + C, 15
API action function, 56
API class, 31, 62, 65 database, 48
API endpoint roundtrip, 54
endpoint, 56, 80
API endpoints, 30
endpoint roundtrip, 54
API, end-points, 63
executing mysql queries, 48
api.js, 47, 62, 65
exit from Node, 15
apt-get update, 34
exit server process, 86
architecture, 46, 54, 56
auth token, 56 frameworks, 4
AuthToken, 56
grant, 39
bash.exe, letsencrypt, 66 grant all privileges, 39
bash.exe, letsencrypt, certbot, 66
bash.ext, 1 home directory, 18
bind-address, 41, 42 HTTPS, https, SSL, 65
building API, 29 HTTPS, https, SSL, pem, pem keys,
secure, 77
certbot, 68
client, 46 IDENTIFIED BY ’PassWord’, 39
conecting to database, 44 identify(), 57
configuring, 11 Installation, 7
CREATE DATABASE installing mysql npm module, 32
database name;, 38 installing mysql server, 32
’localhost’;, 39 installing mysql server on Ubuntu, 34
89
90 INDEX
installing mysql server on Windows, nano control-W, 41

33 nano search, 41
installing node on Mac, 8 node index.js, 19
installing on Windows, 9
on Mac, 8
javascript grammar, 88 on Windows, 9
json(), 58
PHP, 4
launching the server, 19
POST, 59, 60
letsencrypt, certbot, 66, 67
process object, 20
libraries, 4
promise, 56
localhost, 44
Mac, 8 remote access, 40
max connections, 51 request, 62, 65
mime, 28 require mysql, 48
mime types, 28 resolve(), 56
multiple endpoints, 80 respond(), 58
my.cnf, 40 response, 62, 65
MySQL, 47 run from any directory, 13
mysql -u root -p, 36 running a .js file from node, 18
mysql API, 48 running applications, 17
mysql configuration file, 40, 41 running server continuously, 23
mysql connection, 48 running the server, 19
mysql grant, 39
saving mysql .cnf file, 42
mysql grant all privileges, 39
SELECT user FROM mysql.user, 38
MySQL promise, 56
Sequel Pro, 44
MySQL query, 48
server, 46
mysql query, 48
server architecture, 54
mysql remote access, 40
server exit events, 86
mysql table example, 45
serving files, 26, 46
mysql user password, 39
set up mysql server, 33
MySQL user table, 45
setting mysql user password, 39
mysql user table, 45
SHOW DATABASES;, 37
MySQL Workbench, 44
SIGINT, 86, 88
mysqld.cnf, 41
sudo apt-get install mysql-server, 34
nano control-O, 42 sudo apt-get update, 34
INDEX 91
sudo ufw allow mysql, 35 ufw allow mysql, 35

systemctl start mysql, 35 uncomplicated firewall, 35
user, 45
table, 44 user login, 56
ufw, 35 Windows, 9

Nodel

Uploaded by

Copyright:

Available Formats

Nodel

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Nodel

Uploaded by

Copyright:

Available Formats

Title: Node.

Primary purpose of Learning Curve Books publishing company is

Learning Curve Books is trademark of Learning Curve Books, LLC – an in-

0.1 Node.js API – What Is This Book About?

0.1 Node.js API – What Is This Book About? . . . . . . . . . . . .

3.1 Project Home Directory . . . . . . . . . . . . . . . . . . . . . . 18

4 Building The API 29

5 Setting Up MySQL Server 33

6 Adding MySQL to api.js 47

6.1 Asynchronous Back-End Problem . . . . . . . . . . . . . . . . . 49

6.16 Master and Stage Servers . . . . . . . . . . . . . . . . . . . . . 86

Figure 1.1: Check Windows Subsystem for Linux (off by default.)

There is now bash.exe that looks as follows when executed:

Figure 1.2: Running Linux ls (list directory contents) command on Windows.

1.0.1 What is bash.exe?

1.0.2 Ubuntu on Windows?

Figure 1.3: Ubuntu for Windows.

1.1 Migrating From Apache to Node

1.2 Node Libraries and Frameworks

1.3 Running Node Server

2.1.1 apt install nodejs

To get started on Windows, hold Window-key and press R.

To see if node is already installed, type node and hit Enter:

In this case I don’t have node installed, so let’s get started!

2.3 Installing Node on Windows

Figure 2.1: Running: wmic os get osarchitecture

Looks like my computer is 64-bit, just as I thought!

Figure 2.2: URL: https://nodejs.org/en/download/

Figure 2.4: I chose default path C:\Program Files\nodejs

2.4 Configuring Node Server Globally

Now that we’re there, execute node command:

Your cursor will turn to right bracket >

2.5 Enable Node To Run From Any Directory

Figure 2.5: Running: rundll32 sysdm.cpl,EditEnvironmentVariables

Running command rundll32 sysdm.cpl,EditEnvironmentVariables will crack

Figure 2.6: Upper half of Environment Variables window.

Figure 2.7: Upper half of Environment Variables window.

Click OK, then click OK again on underlying window.

Figure 2.8: Now you can run node from anywhere.

Type node then type console.log(”Node Anywhere.”) and watch output.

2.6 Exit from Node

So far we only installed a dormant node.exe on the hard drive.

3.1 Project Home Directory

3.2 Running a file via Node

Figure 3.1: Run: node index.js

3.3 The process Object

Figure 3.2: Printing out contents of process object.

3.4 Adding Packages To index.js

Figure 3.3: Run: node index.js

Figure 3.4: Output: .js

3.5 Next Steps

3.6 Running Application Server Continuously

To continuously listen for connections we need to write server code in index.js.

Figure 3.5: The bare-bones Node server.

There is now a back-end server running from your computer.

3.8 Serving Files Continuously

We’ve just set up our Node server to listen to URL requests.

Update index.js with code above and request 127.0.0.1:3000/index.html or

Building The API

GRANT ALL PRIVILEGES ON . TO ’felix’@’localhost’;