Task Statement 3.3: Determine high-performing database solutions.
📘AWS Certified Solutions Architect – (SAA-C03)
1. What is Database Replication?
Database replication is the process of copying data from one database (primary) to one or more other databases (replicas).
- The primary database handles writes (INSERT, UPDATE, DELETE)
- The replica databases mainly handle reads (SELECT)
This helps improve:
- Performance
- Availability
- Scalability
2. Why Replication is Important
Replication is used in AWS to solve key performance and reliability challenges:
1. Improve Read Performance
- Applications often have more read requests than write requests
- Replicas handle read traffic, reducing load on the primary database
2. Increase Availability
- If the primary database fails, replicas can be promoted (in some cases)
- Helps reduce downtime
3. Disaster Recovery
- Replicas can exist in different Availability Zones or Regions
- Protects against infrastructure failure
4. Offload Workloads
- Reporting, analytics, and backups can run on replicas instead of the primary
3. Types of Replication in AWS
A. Read Replicas
This is the most important concept for the exam.
Key Features:
- Copies data from the primary database
- Used for read-only workloads
- Supports asynchronous replication
Supported Services:
- Amazon RDS (MySQL, PostgreSQL, MariaDB, Oracle, SQL Server)
- Amazon Aurora
How it Works:
- Data is written to the primary database
- Changes are copied to replicas
- Applications read from replicas
Important Characteristics:
- Eventually consistent (not immediate)
- Can have multiple replicas
- Can be in:
- Same Availability Zone
- Different Availability Zones
- Different Regions (cross-region replication)
B. Multi-AZ Deployment (Different from Read Replicas)
This is often confused with read replicas.
Key Features:
- Synchronous replication
- Used for high availability, not scaling reads
- Standby instance is not used for reading
Behavior:
- Primary → synchronous copy → standby
- Automatic failover if primary fails
Key Difference:
| Feature | Read Replica | Multi-AZ |
|---|---|---|
| Purpose | Improve read performance | High availability |
| Read access | Yes | No |
| Replication type | Asynchronous | Synchronous |
| Failover | Manual | Automatic |
C. Aurora Replication
Amazon Aurora has advanced replication features:
Key Features:
- Automatically creates up to 15 read replicas
- Replication is faster than standard RDS
- Uses shared storage across replicas
Benefits:
- Low latency (milliseconds)
- Automatic failover
- High scalability
Aurora Replica vs RDS Replica:
- Aurora replicas share storage → faster
- RDS replicas use separate storage → slower
4. Types of Replication Modes
1. Asynchronous Replication
- Data is copied after it is written to the primary
- Small delay (replication lag)
- Used by:
- Read replicas
Pros:
- Faster performance
- Scales easily
Cons:
- Data may not be up-to-date
2. Synchronous Replication
- Data is written to both primary and replica at the same time
- No data loss
Pros:
- Strong consistency
Cons:
- Higher latency
Used in:
- Multi-AZ deployments
5. Replication Lag
Replication lag = delay between primary and replica data
Causes:
- High write traffic
- Network latency
- Large transactions
Impact:
- Users may read outdated data
Monitoring:
- Use Amazon CloudWatch metrics
6. Use Cases (IT-focused)
1. Web Applications
- Primary handles writes (user data updates)
- Replicas serve read requests (fetching data)
2. Reporting Systems
- Reports run on replicas to avoid affecting production performance
3. Data Warehousing Pipelines
- Data is replicated to another database for analytics processing
4. Global Applications
- Cross-region replicas reduce latency for users in different locations
7. Cross-Region Replication
- Read replicas can be created in another AWS Region
- Useful for:
- Disaster recovery
- Global applications
Benefits:
- Reduced latency for global users
- Regional isolation
8. Promotion of Read Replica
- A read replica can be promoted to standalone database
When used:
- Primary database failure
- Migration scenarios
Important:
- After promotion:
- It becomes independent
- No longer receives updates from the original primary
9. Best Practices for the Exam
1. Use Read Replicas When:
- Application is read-heavy
- Need to scale reads
- Reporting queries are slowing down the database
2. Use Multi-AZ When:
- High availability is required
- Automatic failover is needed
3. Use Aurora When:
- Need high performance and fast replication
- Need many replicas
4. Watch for Replication Lag:
- Avoid using replicas for critical real-time data reads
5. Combine Solutions:
- Multi-AZ + Read Replicas = high availability + scalability
10. Common Exam Questions Patterns
Be prepared for questions like:
Scenario 1:
- Application has heavy read traffic
- Solution → Add read replicas
Scenario 2:
- Need automatic failover
- Solution → Multi-AZ deployment
Scenario 3:
- Global users need low latency reads
- Solution → Cross-region read replicas
Scenario 4:
- Need fast replication and many replicas
- Solution → Amazon Aurora
11. Quick Summary
- Read Replicas → Improve read performance (asynchronous)
- Multi-AZ → High availability (synchronous)
- Aurora → Advanced, fast, scalable replication
- Replication Lag → Delay in data syncing
- Cross-region → Disaster recovery + global access
