Introduction

Let's build an online store! So... we need some items in the store before we can do any business :) we'll implement a Product API based on a given OpenAPI specification. The Product API represents a simple first step e-commerce system with endpoints for creating and retrieving products.

Instructions

Prepare Credentials

Retrieve you temporary credentials from Learner's Lab. Enter your configuration when prompted:

aws configure

And set your session token:

aws configure set aws_session_token <YOUR-TEMP-SESSION-TOKEBN>

Apply Infrastructure

cd terraform
terraform init
terraform apply -auto-approve

Send Requests

Make sure you are in terraform folder

Get public ip address

aws ec2 describe-network-interfaces \
--network-interface-ids $(
    aws ecs describe-tasks \
    --cluster $(terraform output -raw ecs_cluster_name) \
    --tasks $(
        aws ecs list-tasks \
        --cluster $(terraform output -raw ecs_cluster_name) \
        --service-name $(terraform output -raw ecs_service_name) \
        --query 'taskArns[0]' --output text
    ) \
    --query "tasks[0].attachments[0].details[?name=='networkInterfaceId'].value" \
    --output text
) \
--query 'NetworkInterfaces[0].Association.PublicIp' \
--output text

Send some requests:

curl http://<PUBLIC-IP-ADDRESS>:8080/albums

Clean Up

terraform destroy -auto-approve

Product API Documentation

This document demonstrates how to interact with the Product API using curl. Each request includes the expected HTTP response code.

---

API Endpoints Overview

#	Endpoint	Method	Description	Expected Response
1	/products	GET	Get all products	200 OK
2	/products/1	GET	Get a specific product	200 OK
3	/products/999	GET	Get a non-existent product	404 Not Found
4	/products	POST	Create a valid product	201 Created
5	/products	POST	Create product with missing fields	400 Bad Request
6	/products	POST	Create product with invalid price	400 Bad Request
7	/products	POST	Create duplicate product	409 Conflict

---

CURL Examples

( Refer in Screenshots/API-Requests folder for sample response )

# Get all products (200 OK)
curl http://localhost:8080/products

# Get a specific product (200 OK)
curl http://localhost:8080/products/1

# Get a non-existent product (404 Not Found)
curl http://localhost:8080/products/999

# Create a valid product (201 Created)
curl -X POST http://localhost:8080/products \
  -H "Content-Type: application/json" \
  -d '{
    "id": "4",
    "name": "Monitor",
    "description": "4K Ultra HD Monitor",
    "price": 399.99,
    "stock": 15
  }'

# Try to create a product with missing fields (400 Bad Request)
curl -X POST http://localhost:8080/products \
  -H "Content-Type: application/json" \
  -d '{
    "id": "5",
    "description": "Missing name and price"
  }'

# Try to create a product with invalid price (400 Bad Request)
curl -X POST http://localhost:8080/products \
  -H "Content-Type: application/json" \
  -d '{
    "id": "5",
    "name": "Free Product",
    "price": 0
  }'

# Try to create a duplicate product (409 Conflict)
curl -X POST http://localhost:8080/products \
  -H "Content-Type: application/json" \
  -d '{
    "id": "1",
    "name": "Another Laptop",
    "price": 799.99
  }'

HttpUser vs FastHttpUser Performance Comparison

Here’s a clear comparison between HttpUser and FastHttpUser based on the performance data provided.

---

1. Single User Performance

Scenario/User	Request Type	Median (ms)	95th Percentile (ms)	99th Percentile (ms)
HttpUser	GET /products	2.27	4	4
HttpUser	GET /products/id	2.22–2.43	4	4
HttpUser	POST /products	3	3	3
FastHttpUser	GET /products	1.25	2	2
FastHttpUser	GET /products/id	1.26–1.29	2–3	2–3
FastHttpUser	POST /products	1.28	2	2

Observation:
Even with 1 user, FastHttpUser has lower latency than HttpUser

2. 50 Users Performance

Scenario/User	Request Type	Median (ms)	95th Percentile (ms)	99th Percentile (ms)
FastHttpUser	GET /products	1.01	2	6
FastHttpUser	POST /products	1	2	6
FastHttpUser	GET /products/id	1.07–1.09	2	5
HttpUser	GET /products	2	4	10
HttpUser	POST /products	2	4	8
HttpUser	GET /products/id	2	3–4	7–8

Observation: Even with 50 user, FastHttpUser has lower latency than HttpUser

---

---

Key Takeaways

1. FastHttpUser is faster and more efficient than HttpUser, especially under load.
2. HttpUser is simpler and mimics standard Python HTTP requests, which may be more realistic for some use cases.

---