Recommending appropriate compute, storage, networking, and databasetechnologies based on requirements

Task Statement 2.1: Design scalable and loosely coupled architectures.

📘AWS Certified Solutions Architect – (SAA-C03)


1. Understanding Requirements (VERY IMPORTANT)

Before choosing any service, always identify:

1. Workload Type

  • Web application
  • Batch processing
  • Real-time processing
  • Analytics workload

2. Key Factors

You must evaluate:

  • Scalability → Does traffic change frequently?
  • Availability → Must it always be online?
  • Performance → Fast response required?
  • Cost → Minimize cost?
  • Management → Fully managed or self-managed?
  • Durability → How important is data safety?

👉 In the exam, questions often hide the answer inside these requirements.


2. Compute Services (Processing Power)

Compute services run your application code.

Main AWS Compute Options

1. Amazon EC2 (Virtual Servers)

Use when:

  • Full control over OS is required
  • Custom software installation is needed
  • Long-running applications

Key features:

  • Flexible instance types
  • Auto Scaling support
  • Works with load balancers

Best for:

  • Traditional applications
  • Lift-and-shift migrations

2. AWS Lambda (Serverless Compute)

Use when:

  • Event-driven workloads
  • No server management required
  • Automatic scaling needed

Key features:

  • Runs code without servers
  • Pay per execution
  • Scales instantly

Best for:

  • API backends
  • Event processing (e.g., file uploads)

3. Amazon ECS / EKS (Containers)

Use when:

  • Applications are containerized
  • Need microservices architecture

Difference:

  • ECS → Easier, AWS-managed
  • EKS → Kubernetes-based, more control

Best for:

  • Scalable microservices
  • DevOps-heavy environments

4. AWS Fargate

Use when:

  • Running containers without managing servers

Best for:

  • Serverless containers

5. AWS Elastic Beanstalk

Use when:

  • Want easy deployment without managing infrastructure

Best for:

  • Beginners
  • Web apps

Compute Decision Tips (Exam)

RequirementBest Choice
Full controlEC2
No server managementLambda
ContainersECS / EKS
Serverless containersFargate
Quick deploymentElastic Beanstalk

3. Storage Services (Data Storage)

Storage depends on how data is accessed and used.

Main Storage Types

1. Amazon S3 (Object Storage)

Use when:

  • Storing files, images, backups
  • Static website hosting

Key features:

  • Highly durable (11 9’s durability)
  • Scalable
  • Multiple storage classes

Storage classes:

  • Standard → frequent access
  • IA → infrequent access
  • Glacier → archival

2. Amazon EBS (Block Storage)

Use when:

  • Attached to EC2
  • Needs low-latency access

Best for:

  • Databases
  • OS disks

3. Amazon EFS (File Storage)

Use when:

  • Multiple EC2 instances need shared access

Best for:

  • Shared file systems
  • Linux workloads

4. Instance Store

Use when:

  • Temporary storage
  • Data loss acceptable

5. AWS Backup / Glacier

Use when:

  • Long-term backup and archive

Storage Decision Tips (Exam)

RequirementBest Choice
File storage (shared)EFS
Block storageEBS
Object storageS3
Temporary storageInstance Store
ArchiveGlacier

4. Database Services

Choosing the correct database is very important in the exam.


1. Relational Databases (SQL)

Amazon RDS

Use when:

  • Structured data
  • ACID compliance needed

Supported engines:

  • MySQL, PostgreSQL, MariaDB, Oracle, SQL Server

Features:

  • Automated backups
  • Multi-AZ for high availability
  • Read replicas for scaling

Amazon Aurora

Use when:

  • Need high performance
  • Cloud-optimized database

Benefits:

  • Faster than standard RDS
  • Auto scaling storage

2. NoSQL Database

Amazon DynamoDB

Use when:

  • Massive scale
  • Low latency (milliseconds)
  • Key-value or document data

Features:

  • Fully managed
  • Auto scaling
  • Serverless

3. In-Memory Database

Amazon ElastiCache

Use when:

  • Need fast caching
  • Reduce database load

Engines:

  • Redis
  • Memcached

4. Data Warehouse

Amazon Redshift

Use when:

  • Analytics
  • Large-scale reporting

5. Specialized Databases

ServiceUse Case
NeptuneGraph data
DocumentDBJSON/document data
TimestreamTime-series data

Database Decision Tips (Exam)

RequirementBest Choice
Structured dataRDS
High performance relationalAurora
NoSQL, scalableDynamoDB
CachingElastiCache
AnalyticsRedshift

5. Networking Services

Networking controls how systems communicate securely and efficiently.


Core Networking Services

1. Amazon VPC

  • Isolated network environment
  • Define IP ranges, subnets

2. Subnets

  • Public subnet → internet access
  • Private subnet → internal systems

3. Internet Gateway

  • Allows internet access

4. NAT Gateway

  • Allows private resources to access internet

5. Elastic Load Balancer (ELB)

Distributes traffic:

Types:

  • ALB → HTTP/HTTPS
  • NLB → TCP/UDP (high performance)
  • GWLB → security appliances

6. Route 53

  • DNS service
  • Routing policies (latency, failover, etc.)

7. CloudFront

  • Content Delivery Network (CDN)
  • Reduces latency

8. Security Groups & NACLs

  • Security Groups → instance-level firewall
  • NACL → subnet-level firewall

Networking Decision Tips (Exam)

RequirementBest Choice
Global DNSRoute 53
Load balancingELB
Private networkingVPC
Faster content deliveryCloudFront
Secure accessSecurity Groups

6. Putting It All Together (Architecture Thinking)

In the exam, you will often combine services:

Example Pattern (Typical Exam Scenario)

  • Frontend → CloudFront + S3
  • Backend → Lambda or EC2
  • Database → DynamoDB or RDS
  • Load Balancer → ALB
  • Networking → VPC with public/private subnets

7. Key Exam Tips

1. Always Match Requirements

  • Don’t choose services randomly
  • Match keywords in the question

2. Look for Keywords

KeywordMeaning
“Fully managed”Avoid EC2
“Serverless”Use Lambda/DynamoDB
“Low latency”DynamoDB / ElastiCache
“Shared storage”EFS
“High throughput”S3 or EBS

3. High Availability

  • Use Multi-AZ
  • Use Load Balancers
  • Use Auto Scaling

4. Cost Optimization

  • Use serverless when possible
  • Use correct storage class

5. Scalability

  • Use Auto Scaling
  • Use managed services

8. Quick Summary (Memory Section)

  • Compute → EC2, Lambda, ECS
  • Storage → S3, EBS, EFS
  • Database → RDS, DynamoDB, Aurora
  • Networking → VPC, ELB, Route 53

Final Advice for Exam

  • Read the question carefully
  • Identify requirements first
  • Eliminate wrong options
  • Choose the simplest and most scalable solution
Buy Me a Coffee