Click here to see the "buggy" versions 🐛

The Scenario

TinyHat.Me is an up and coming startup that provides an API to allow users to try on tiny hats via a REST API. The startup is looking to you to ship some awesome new features! While we have a frontend working to serve tiny hats from a database, the startup is looking to add some exciting new features. Today, as the newest backend engineer at TinyHat.Me, you are going to be learning how the microservices work, as well as fix a few outstanding bugs from our customers!

Featured In

• New Relic One and Pixie AWS Workshop
• New Relic Pixie Tutorial

Run tinyhats Locally without S3

Download minikube

Follow instructions here to install minikube.

Deploy Cluster

minikube start

Download Kubernetes Manifests

Clone this repository and cd into root directory.

kubectl apply -f kube-local

Test Cluster

To test the API directly:

minikube service -n default --url gateway-service

To interact with the API through the frontend:

minikube service -n default --url frontend-service

Cleaning Up

minikube delete

Deploying tinyhats with EKS and S3

S3 Bucket

Click here to access the S3 console.

Creating the bucket

Click Create bucket.

1. Name your bucket in the Bucket name field. (Example: tinyhats)
2. Edit the Public Access settings to match the configuration below:

3. Leave all other settings as defult and click Create bucket.

Allowing public access

On the S3 console, click on the newly created S3 bucket by identifying it with the name you assigned it to. (Example: tinyhats)

1. Click on Permissions and scroll down to Bucket policy.
2. Click Edit and paste the below policy in the editor, remembering to replace tinyhats with your bucket name

{
    "Version": "2012-10-17",
    "Statement": [
        {
            "Sid": "AddPerm",
            "Effect": "Allow",
            "Principal": "*",
            "Action": "s3:GetObject",
            "Resource": "arn:aws:s3:::tinyhats/*"
        }
    ]
}

3. Click Save changes. If you configured your bucket correctly, you should see the labels Publicly accessible and Access labeled as Public.

Deploying the Cluster

Ensure you have a running cluster and kubectl configured. Obtain your AWS ID and Secret. Export these into environment variables.

export S3_ID=[your S3 ID]
export S3_SECRET=[your S3 secret]

Clone this directory. Change directory (cd) into the kube folder and run the below command to apply your S3 secrets and deploy the Kubernetes resources.

for f in *.yaml; do envsubst < $f | kubectl apply -f -; done

Using kubectl get pods --watch, monitor when the pods are finished deploying. Use kubectl get services to copy the URI of the gateway-service or frontend-service for testing.

API Documentation

Parameters

api.tinyhat.me?number=2 ?number (optional): How many hats you want!

[GET] /

api.tinyhat.me Returns a random tinyhat on Bob Ross.

[GET] /${hat style}

api.tinyhat.me/santa Returns a specific hat style on Bob Ross.

[GET] /api/hats

api.tinyhat.me/api/hats Returns a JSON object of all hats' image links and style identifiers.

[POST] /

api.tinyhat.me POST with multipart/form-data: image with face

Returns a random tinyhat on your POSTed image.

[POST] /${hat style}

api.tinyhat.me/santa POST with multipart/form-data: image with face

Returns a specific hat style on your POSTed image.

[POST] /add

api.tinyhat.me/add POST with multipart/form-data: image with hat and name attribute

Adds the image to the admin page to queue for approval.

Flowchart