Bootstrapping Microservices with AWS - Part 1

I recently completed a batch at the Recurse Center and got to go on a self-directed learning journey through so many topics! During my second week I quickly made some friends who suggested that I join their Bootstrapping Microservices group and I was excited to join since I had always wanted to have a hands on dive into understanding microservice architecture with Docker and Kubernetes. During our group meetings we worked through the Bootstrapping Microservices textbook by Ashley Davis together.

The textbook uses Azure as a cloud provider but since I had experience with AWS I wanted to challenge myself to deploy all of the example code using AWS as my cloud provider. It was a fun experience tweaking the processes and the code for AWS and I bet there’s others out there who’d like to go through the book with AWS so I thought it'd be nice to write a guide for how I completed all the chapters in the book using AWS as my cloud provider.

This guide will be organized according to the chapters in the book. I’ll only write entries for when the instructions in the book differ from what you need to do with AWS. The first two chapters just cover some theory and running the app locally so I’ll skip those and dive right into chapter 3!

A couple quick notes before we begin: to simplify the commands, I’m using us-east-2 as my region for everything. Make sure to pick a region and stick with it. If you use a region other than us-east-2, remember to replace us-east-2 with your region for all of your commands.

I’ll also be using the CLI which you’ll need to have credentials for. It’s not recommended to generate credentials using your root user. Ideally, you’ll generate temporary credentials but I find that to be a little unwieldy. Instead, I used my root account to create an IAM user with admin permissions. I then created an access key with my new IAM user and copied the secret access key into my password manager. As the console warns you, be careful with these credentials! Never include these credentials in a git repo and be sure to store them somewhere secure.

Chapter 3

3.9.1 Creating a private container registry

At the end of chapter 3 you get to deploy a Docker image to a private container registry. AWS’s private container registry is ECR. There’s two ways you can create a registry, through the console and through the AWS CLI.

To deploy through the console, first navigate to ECR in the AWS console. You can use the search bar at the top of the console. Once you get to the create repository page, you can fill at the details as shown in the photo. The textbook (using Azure) mentions that your repository name will need to be globally unique, but that’s not true for AWS as they add your account number as a subdomain of your repository URL. I just used “video-streaming” and you can too!

Make sure to check what region you’re in as you deploy too as you’ll need to remember this for later. You can change it by selecting a new region from the dropdown menu in the top right.

To deploy through the CLI, make sure your aws credentials are set in your .aws directory by running aws configure and run the following command

aws ecr create-repository --repository-name video-streaming --region us-east-2

If this command is successful you should get some JSON back that gives you some info about your new repository. Copy your repository URI and save it for later. It should look something like this

YOUR_ACCOUNT_ID.dkr.ecr.us-east-2.amazonaws.com/video-streaming

3.9.2 Pushing our micro service to the registry

After you create your registry, it’s time to push! This step must be done using the CLI.

First, you’ll need to receive an authentication token to authenticate your Docker client. This step is pretty different than as outlined in the book with Azure. You can use the following command to do so, just be sure to use your account ID and region!

aws ecr get-login-password --region us-east-2 | docker login --username AWS --password-stdin YOUR_ACCOUNT_ID.dkr.ecr.us-east-2.amazonaws.com

Yes, the username is AWS. That was a little confusing to me at first too. This command should give you the message “Login Succeeded”

You should have already built your Docker image with the instructions in the textbook. The textbook should help you with tagging your image with the docker tag command. Instead of the Azure registry URL, your registry should just look something like this

YOUR_ACCOUNT_ID.dkr.ecr.us-east-2.amazonaws.com/video-streaming:latest

Pushing your image should similarly be easy to figure out with the textbook. Just run

docker push YOUR_ACCOUNT_ID.dkr.ecr.us-east-2.amazonaws.com/video-streaming:latest

You now have a private ECR repo with your built Docker image!

3.9.4 Deleting your container registry

Deleting your registry if you choose to do so is easy! You can simply delete it in the console or use the aws ecr delete-registry —registry-name video-streaming command. But since the main motivation to do so is saving money, let’s talk about pricing.

The textbook uses Azure in large part because it comes with free credits. AWS does not, but it does come with something similar: the Free Tier. The Free Tier can be a little confusing since there are actually three different offerings within the Free Tier: a 12-month Free Tier, the Always Free tier and short term trials. Luckily, AWS provides a pretty easy to understand webpage to figure out what’s free for what service.

The Free Tier for ECR has a 50GB limit and falls under the 12-month Free Tier. This starts from when your AWS root account was created. If you’re just starting out with AWS and your AWS account is less than a year old, storing images from the book will be free until a year from your account creation as they will easily fall within the 50 GB free tier. But if you’re like me and your AWS account is a year old or older, storing containers on ECR will cost you money. Don’t worry though, as the cost is very low! At 10¢/GB/month and with an approximate container size of 400MB, it will only cost you 4¢ a month to store your container for this chapter.

I highly recommend setting up a cost alert while working through this book! It’s saved me on multiple occasions. I set my initial alert to be at $1 per month and to notify me if it ever reaches 80% of that limit.

That’s all for now! Stay tuned for Chapter 4!