Task Statement 2.1: Design scalable and loosely coupled architectures.

📘AWS Certified Solutions Architect – (SAA-C03)

1. What is Caching?

Caching means storing frequently accessed data in a temporary, high-speed storage layer so that future requests for that data can be served faster.

Instead of repeatedly fetching data from a slow system (like a database or external API), the system retrieves it from a faster layer called a cache.

In AWS architecture, caching is very important for:

Improving performance
Reducing latency
Lowering database load
Reducing cost
Increasing scalability

For the SAA-C03 exam, you must understand:

Where caching can be implemented
Which AWS services provide caching
When to use each caching strategy
How caching improves scalability and loose coupling

2. Why Caching Is Important for Scalable Architectures

When a system grows:

More users send requests
Databases get overloaded
Applications slow down
Costs increase

Caching helps by:

✅ Reducing the number of direct database calls
✅ Reducing compute usage
✅ Reducing network calls
✅ Improving response time
✅ Supporting horizontal scaling

In scalable systems, caching reduces dependency on backend systems, which helps create loosely coupled architectures.

3. Types of Caching in AWS Architectures

For the SAA-C03 exam, caching can be divided into:

Client-side caching
CDN caching (edge caching)
Application-level caching
Database caching
DNS caching

Let’s understand each one clearly.

4. Edge Caching (CDN Caching)

Service: Amazon CloudFront

What It Does

CloudFront stores copies of static and dynamic content at edge locations around the world.

When a user requests content:

The request is served from the nearest edge location
The origin server (like S3 or EC2) is not contacted every time

What Can Be Cached?

Images
Videos
CSS, JavaScript files
API responses
HTML pages

Benefits

Very low latency
Reduced load on origin servers
Global performance improvement
DDoS protection integration

Exam Tip

If the question mentions:

Global users
Static content
Reducing latency worldwide
Reducing load on origin

→ The answer is usually CloudFront caching.

5. In-Memory Caching (Application-Level Caching)

Service: Amazon ElastiCache

What It Does

ElastiCache stores data in memory (RAM) for extremely fast access.

It supports:

Redis
Memcached

Memory access is much faster than database disk access.

When Is It Used?

Frequently accessed database query results
Session data
Leaderboards
Frequently used configuration data
API response caching

How It Improves Scalability

Instead of:
Application → Database (every time)

It becomes:
Application → Cache → Database (only if needed)

This reduces database load and allows the system to handle more users.

Redis vs Memcached (Exam Important)

Redis

Supports replication
Supports persistence
Supports complex data types
Multi-AZ support

Memcached

Simple key-value
No replication
No persistence
Multi-threaded

Exam Tip:

If high availability or replication is required → Choose Redis.

6. Database Caching

Service: Amazon RDS

RDS does not automatically cache everything, but you can:

Use read replicas
Use ElastiCache with RDS
Enable query caching (engine-dependent)

Read Replicas

Used for:

Offloading read traffic
Improving scalability

If the question says:

Heavy read workload
Reduce primary DB load
Improve performance

→ Use RDS Read Replica

7. DynamoDB Caching

Service: Amazon DynamoDB

Feature: DynamoDB Accelerator (DAX)

What is DAX?

DAX is an in-memory cache for DynamoDB.

It:

Reduces read latency from milliseconds to microseconds
Is fully managed
Works as a caching layer in front of DynamoDB

Exam Tip

If the question says:

DynamoDB performance issue
Improve read performance
Reduce latency
No application change required

→ The answer is DAX

8. API Caching

Service: Amazon API Gateway

API Gateway supports built-in caching.

When enabled:

Responses are cached
Repeated API calls return cached data
Backend load decreases

Used For:

Frequently requested API responses
Public APIs
Microservices architecture

9. Client-Side Caching

In this strategy:

The client (browser or application) stores responses locally
Uses HTTP cache headers

Example:

Cache-Control
Expires
ETag

Reduces repeated calls to backend services.

Exam usually tests:

How to reduce API calls
Improve performance
Reduce cost

10. TTL (Time to Live)

TTL defines how long data remains in cache.

After TTL expires:

Cache invalidates
Data must be fetched from origin again

Choosing TTL is very important:

Short TTL:

More fresh data
More backend calls

Long TTL:

Less backend load
Risk of stale data

11. Cache Invalidation

Sometimes cached data must be removed manually.

Methods:

Automatic TTL expiry
Manual invalidation
Cache versioning

For CloudFront:
You can invalidate specific files.

12. Write Strategies in Caching (Very Important for Exam)

There are 3 main write patterns:

1. Cache-Aside (Lazy Loading)

Flow:

Application checks cache
If not found → fetch from database
Store in cache
Return to user

Most common pattern.

Used in:

ElastiCache
DAX

2. Write-Through

Flow:

Application writes to cache
Cache writes to database

Ensures:

Cache always updated
Data consistency improved

3. Write-Behind (Write-Back)

Flow:

Write to cache
Cache updates database later

Improves performance but:

Risk of data loss if cache fails

Exam may test:

Performance vs consistency trade-offs

13. How Caching Supports Loose Coupling

Loose coupling means:
Components do not depend tightly on each other.

Caching helps because:

Application does not always depend on database
Backend systems can fail without immediate impact
Traffic spikes are absorbed by cache
Reduces direct system-to-system dependency

This improves:

Fault tolerance
Scalability
Availability

14. When NOT to Use Caching

Caching is not good for:

Frequently changing data
Real-time financial data
Critical real-time updates
Very small workloads

Exam may test this by asking:
“Which solution ensures the most up-to-date data?”

Answer:
Avoid heavy caching.

15. Cost Optimization Using Caching

Caching reduces:

RDS compute cost
DynamoDB read capacity usage
EC2 processing load
API calls
Data transfer

CloudFront reduces:

Origin data transfer cost

16. Common Exam Scenarios and What to Choose

Scenario	Best Service
Global content delivery	CloudFront
Reduce RDS read load	Read Replica or ElastiCache
Improve DynamoDB read performance	DAX
Cache API responses	API Gateway caching
Store session data	ElastiCache Redis
Reduce database cost	Add caching layer

17. Architecture Pattern for Scalable Design

Typical scalable architecture:

Users → CloudFront → Load Balancer → EC2 → ElastiCache → RDS

Caching reduces pressure on:

EC2
Database
Network

This allows:

Horizontal scaling
Auto Scaling groups
Microservices growth

18. Key Differences to Remember for Exam

CloudFront = Edge caching
ElastiCache = In-memory caching
DAX = DynamoDB caching
Read Replicas = Database scaling
API Gateway caching = API response caching

19. High Availability and Caching

ElastiCache Redis:

Multi-AZ
Automatic failover

CloudFront:

Global redundancy

DAX:

Multi-node cluster support

20. Final Exam Checklist (Very Important)

You must understand:

✔ When to use CloudFront vs ElastiCache
✔ Redis vs Memcached differences
✔ DAX purpose
✔ TTL and invalidation
✔ Write-through vs Cache-aside
✔ How caching reduces database load
✔ How caching improves scalability
✔ How caching supports loose coupling

Final Summary

Caching is one of the most important strategies in scalable AWS architectures.

It:

Improves speed
Reduces backend pressure
Reduces cost
Improves availability
Supports loosely coupled design

For SAA-C03, expect scenario-based questions like:

“Application performance is slow due to high read traffic.”
“Users globally experience high latency.”
“Database is overloaded with read queries.”
“Need microsecond latency for DynamoDB.”

In those cases, caching is usually the correct architectural improvement.