Introduction to Our Real Life Application

In this hands-on lab, you will deploy a microservices-based pet adoption website on AWS and run Fault Injection Simulator (FIS) experiments. The application showcases pets available for adoption, complete with images, pricing, and details, and leverages AWS services for high availability, scalability, and resilience.

Note

This tutorial uses real AWS resources and may incur charges. Clean up your environment after completion to avoid unexpected costs.

The complete source code for each microservice is available on GitHub. This lab is adapted from the AWS Workshop Studio and provides step-by-step instructions to replicate the environment in your AWS account.

The pet adoption system follows microservices principles and is composed of three core services. Each service uses multiple AWS components to ensure redundancy, scalability, and secure data handling.

Microservice	Compute	High Availability	Auto Scaling	Core Services
Pet Adoption Web	Amazon EKS (2 EC2 worker nodes)	Multi-AZ EC2 spread	EC2 Auto Scaling Group	Application Load Balancer (ALB)
Pet Search API	Amazon ECS (2 EC2 instances)	Multi-AZ EC2 spread	EC2 Auto Scaling Group	ALB, DynamoDB (metadata), Amazon S3 (images)
Pet Payment API	AWS Fargate (2 tasks)	ALB balanced tasks	Fargate Auto Scaling	API Gateway, AWS Lambda, Amazon Aurora PostgreSQL

1. Pet Adoption Web Microservice

Compute: Runs on an Amazon EKS cluster backed by two EC2 worker nodes.
High Availability: EC2 instances are distributed across two separate Availability Zones.
Traffic Distribution: Incoming web traffic is routed through an Application Load Balancer (ALB).
Scalability: The EC2 nodes are managed by an Auto Scaling group that scales out based on CPU and network utilization.

2. Pet Search API Microservice

Compute: Deployed in an Amazon ECS cluster on two EC2 instances.
High Availability: Instances run across two Availability Zones for fault tolerance.
Traffic Distribution: An ALB fronts the ECS service to balance search requests.
Scalability: The ECS cluster uses an Auto Scaling group to adjust capacity dynamically.
Data Storage:
- Amazon DynamoDB stores pet metadata (name, price, availability).
- Amazon S3 hosts pet images, enabling efficient content delivery.

3. Pet Payment API Microservice

Compute: Leverages AWS Fargate to run two serverless containers.
High Availability: Fargate tasks are spread across multiple Availability Zones and balanced by an ALB.
Scalability: Task counts adjust automatically via AWS Application Auto Scaling.
API Integration:
1. Amazon API Gateway receives payment requests.
2. AWS Lambda functions validate the request, fetch pet details from DynamoDB, and process images from S3.
Database:
- Amazon Aurora PostgreSQL handles transactional data with a Multi-AZ configuration (one writer, one reader) for automatic failover.

The image is a diagram of a cloud architecture for a pet adoption application, showing components like a Virtual Private Cloud (VPC), availability zones, and various AWS services such as Lambda, S3, and databases. It illustrates the flow of data and services between users and the cloud infrastructure.

As illustrated above, the architecture spans multiple Availability Zones, incorporates load balancers and auto-scaling groups, and is designed to maximize uptime and resilience.

Links and References

Watch Video

Watch video content