Summary of AWS Services:

General Services:

1. CodePipeline
Builds code and deploys it to various AWS services. Charged per pipeline. Can be used with
Elastic Beanstalk for zero-downtime deployments.

2. Elastic Load Balancing (ELB)

Distributes traffic across servers like EC2, Lambda, or Fargate. Charges are based on running
hours and the amount of traffic handled.

3. RDS (Relational Database Service)

Relational database hosting, charged similarly to EC2 (based on resources). Requires a restart
to resize the server.

4. Route 53
Domain and DNS management service.

5. S3 (Simple Storage Service)

Cloud storage for objects. Charges depend on data stored, storage type, and retrieval costs.

Application Executors:

1. EC2 (Elastic Compute Cloud)

Scalable virtual servers. Charges are based on resources (RAM, CPU, storage) and running
time.

2. Fargate
Runs containerized applications like Docker. Charges depend on the resources allocated and
runtime duration.

3. Lambda
Executes code only when triggered (e.g., by HTTP request or S3). Charges are based on
memory usage, execution time, and number of requests.

4. Lightsail
Simple virtual machines like EC2. Resources are fixed; resizing requires cloning the machine
and restarting.

5. Elastic Beanstalk
Integrates EC2, RDS, and ELB for easy configuration and deployment. Supports autoscaling
and multiple languages (Python, NodeJS, Java, Go). Billing combines the costs of the
resources used.

Availability Zones (AZ):

Services can be deployed across multiple AZs for redundancy, increasing costs proportionally.

For cost estimation, the AWS Pricing Calculator (AWS Calculator) can be used to plan and predict
expenses effectively.
Summary of the Client Email:

Sender: Lilly Sawyer, CTO, Fastier

Subject: AWS Services for Web Application Hosting

Key Details from the Email:

 Company Introduction: Fastier is an online tool that helps organize things quickly.

 Current Problems:

1. Slow response times during peak traffic.

2. Server crashes due to running out of memory.

3. Downtime during deployments.

4. Lack of a disaster recovery plan.

 Existing Setup:

o Single AWS EC2 instance (t3.medium, 4GB RAM).

o SPA React frontend, Python + Flask backend.

o PostgreSQL database.

 Requirements:

o A new architecture to resolve current issues.

o Budget is flexible but not unlimited.

o Anticipated steady linear growth in the future.

Request: Provide a solution using AWS services to address their hosting challenges.

SUMMARY –

Summary of the Architecture Plan for Lilly:

Key Considerations:

 Startup Flexibility: Fastier can likely make changes to its application stack and migrate data
into AWS with minimal hurdles.

 Current Pain Points: Addressing downtime, scaling issues, and performance lags is critical.

 Static Content Hosting: Hosting static content for the mixed API and web application is
necessary.

Proposed Solution:

 Elastic Beanstalk:

o Supports Python applications, making it compatible with their stack.

o Provides automatic scaling to handle increased traffic.

 o Includes Blue/Green deployments to eliminate downtime during updates.

Services in the Solution:

1. Route 53: Manages domain and DNS for the application.

2. Elastic Load Balancing: Distributes traffic across multiple servers, improving reliability and
handling surges.

3. Elastic Beanstalk with Autoscaling EC2 Group: Automates the deployment and scaling of
application instances.

4. RDS (PostgreSQL): A managed database service to handle relational data with enhanced
scalability.

5. S3: Stores static content efficiently and cost-effectively.

6. CodePipeline: Automates application deployment for quicker updates and streamlined

workflows.

7. Additional Availability Zone: Adds redundancy to improve disaster recovery capabilities.

Why Elastic Beanstalk?

 Simplifies scaling and deployment, reducing operational overhead.

 Handles automatic scaling and infrastructure management, unlike EC2 or Lightsail.

 Alternatives like Lambda or Fargate are viable but more complex to set up.

Cost Considerations:

 Variable Costs: Costs for Elastic Beanstalk, RDS, and Load Balancing depend on traffic and
resource usage.

 Storage Costs: S3 is priced based on the amount of data stored and accessed.

 Fixed Costs: Route 53 and CodePipeline have more predictable charges.

 Scaling: Autoscaling adjusts server usage dynamically, reducing unnecessary expenses

during off-peak hours.

This architecture provides a scalable, reliable, and cost-efficient solution tailored to Fastier’s
growth and application needs.

ANSWER OF

Subject: AWS Architecture Solution for Fastier Web Application Hosting

Dear Lilly,

Thank you for reaching out and sharing the challenges Fastier is facing with its current hosting
setup. I understand how critical it is to ensure your web application performs reliably, scales
seamlessly during peak traffic, and maintains high availability. Based on your requirements and
our expertise, we’ve designed an AWS architecture tailored to address your challenges and
support your projected growth.

Below, I’ve detailed the proposed architecture, including the components, reasons for their
selection, and an overview of cost considerations.

Proposed Architecture

1. Route 53

AWS Route 53 will handle your domain and DNS management. It ensures highly available and
scalable routing of user requests to the appropriate resources. Using Route 53, we can integrate
health checks to automatically route traffic away from unhealthy instances, further improving
reliability.

Why chosen: Provides robust DNS management and supports health checks. Seamlessly integrates
with other AWS services like Elastic Load Balancing.

Costs: Billed based on the number of hosted zones and DNS queries. Monthly costs may vary
depending on the volume of user traffic.

2. Elastic Load Balancing (ELB)

ELB will distribute incoming traffic across multiple EC2 instances within your Elastic Beanstalk
environment. It ensures that no single instance is overwhelmed and supports seamless scaling
during high-traffic periods.

Why chosen: Prevents server crashes by balancing traffic and ensures high availability. Supports
health checks to route traffic only to healthy instances.

Costs: Charged based on running hours and data processed. Costs may increase during peak traffic
due to higher data processing.

3. Elastic Beanstalk with Autoscaling EC2 Group

Elastic Beanstalk simplifies the deployment and management of your Python and Flask
application. It will automatically manage load balancing, scaling, and monitoring. Autoscaling
ensures your application scales up during high traffic and down during low usage, optimizing
resource usage and costs.

Why chosen: Built-in support for Python applications, automatic scaling, and Blue/Green
deployment for zero-downtime updates.

Costs: Charges are based on the EC2 instances used (e.g., type, size, and duration). Autoscaling
helps manage costs by adjusting resources dynamically.

4. RDS (Relational Database Service)

RDS will host your PostgreSQL database, offering managed backups, updates, and scaling. You can
configure it for Multi-AZ deployment to improve fault tolerance and availability.

Why chosen: Provides a scalable, reliable, and managed database solution that reduces
operational overhead.

Costs: Charged based on instance type, storage, and optional Multi-AZ deployments. Costs may
increase with larger datasets and higher read/write demands.

5. S3 (Simple Storage Service)

S3 will store static assets for your React SPA (e.g., images, scripts, stylesheets) and provide a highly
available, cost-effective solution. You can also configure it for versioning and backups.

Why chosen: Ideal for storing and serving static content, with low latency and high durability.

Costs: Charges depend on the amount of storage used, data retrievals, and data transfer out. Costs
will vary with the volume of stored and retrieved assets.

6. CodePipeline

CodePipeline will automate your CI/CD pipeline, ensuring smooth and efficient deployment of
application updates. This will integrate seamlessly with Elastic Beanstalk for Blue/Green
deployments to minimize downtime.

Why chosen: Automates deployment, reducing manual effort and deployment-related downtime.

Costs: Billed per active pipeline. Costs will depend on the number of pipelines and deployment
frequency.

7. Additional Availability Zone (Optional)

For added redundancy and disaster recovery, we recommend deploying resources across multiple
availability zones. This ensures your application remains accessible even if one zone experiences an
outage.

Why chosen: Enhances fault tolerance and improves application uptime.

Costs: Costs for services like EC2, RDS, and ELB will double for resources deployed across multiple
zones.

Cost Overview

While we cannot provide exact figures without more data, here is a high-level overview:

 Variable Costs: Traffic-dependent services like Route 53, ELB, and S3 will fluctuate based on
user activity.

 Fixed Costs: Base resources like EC2 instances, RDS, and CodePipeline will incur predictable
monthly charges.
  Scalability: Autoscaling and usage-based billing for services like Elastic Beanstalk and RDS
will help optimize costs based on your actual needs.

For detailed cost estimation, I recommend using the AWS Pricing Calculator to simulate your
monthly expenses based on projected usage.

Next Steps

1. We can arrange a follow-up call to discuss this solution and address any questions or
concerns.

2. If you’re ready, we can begin with a proof of concept to implement this architecture
incrementally.

3. Alternatively, we can explore other AWS services like Lambda or Fargate, depending on
your long-term goals and application needs.

We’re excited to support Fastier’s journey and ensure your application performs seamlessly as you
continue to grow.

Best regards,
[Your Name]
Solutions Architect
AWS