While creating your SAAS on AWS, your need a basic networking stack. This how-to guide provides you terraform instructions to achieve that.

AWS is a very big and complex platform. Creating a SAAS on AWS can be a daunting task, especially if you are not familiar with AWS. There are so many configurations about VPC, subnets, security groups, etc to be created before you even begin provisioning your application/database servers. The UI and documentation can be very confusing. 

So I prefer a script/code-based approach to creating infrastructure.&nbsp;

 Terraform is a popular choice to create your AWS infrastructure programmatically.&nbsp;This how-to guide helps you create a basic networking stack using terraform that you can use to create your SAAS.&nbsp;

If you are starting a new SAAS project and planning to run your infrastructure on AWS, then you are going to need&nbsp;

The overall setup will look like below -&nbsp;

If you are using mac, you can get terraform using homebrew easily.&nbsp;

The entire source code is available here&nbsp;

Let's create a skeleton for our scripts. Networking setup tends to grow over time with more security groups, firewall rules, etc. Hence it is ideal to create a separate networking module instead of putting everything in one main script.

Now, if I print the tree, this is how it looks

As your project grows, the size of scripts will grow. Hence, it is ideal to create a module.

This is the most important part of the project.&nbsp;

The highest level resource we create is a VPC.&nbsp;

Each VPC has 1 private and 1 public subnet in each availability zone.

An important thing to notice here is&nbsp;&nbsp;<code class="inline-code">map_public_ip_on_launch</code> is true for public and false for the private subnet.

We also additionally need an elastic IP for our NAT.

The IDs of the VPC, subnets, and the security group that the module creates will be often required when you create your other AWS resources like ec2 instances or ECS cluster, etc.&nbsp;

That is why its important to export these values from the terraform module.

Once the networking module is defined, we want to invoke that module from the main script and pass the necessary variables.&nbsp;

To make our stack configurable, we also need to define the required parameters are the input to the main script.&nbsp;

To create infrastructure, we need initial information like - region, zones, CIDR blocks, etc. We can define these are variables as a good practice.

Now, we need to provide the actual values. While we can pass these from the command line, an easier way is to define a <code class="inline-code">terraform.tfvars</code> file.&nbsp;

You can learn more about CIDR blocks <a href="https://www.digitalocean.com/community/tutorials/understanding-ip-addresses-subnets-and-cidr-notation-for-networking">here</a>.

To invoke the script and you need first setup your AWS account access.&nbsp;&nbsp;

Terraform programmatically creates infrastructure for you. To do so, it needs your AWS access keys.&nbsp;You can get your access keys from your <u class="cdx-underline">AWS console &gt; IAM &gt; User &gt; Security Credentials</u> section.

A better way to pass these credentials is to set them in below environment variables.

To create the actual stack, just run the usual terraform scripts. As mentioned before, it will take your input parameters from the terraform.tfvars files.

Creating a proper network stack for your SAAS on AWS is not trivial. There are a lot of resources that need to be provisioned.&nbsp;

Scripts shared in this how-to guide help you get a basic stack that you can later build upon iteratively.

How to create a VPC, subnet, and networking setup in AWS using terraform?

Nilesh Mahajan