Storage options and characteristics (for example, durability, replication)

Task Statement 2.2: Design highly available and/or fault-tolerant architectures.

📘AWS Certified Solutions Architect – (SAA-C03)


1. What Are Storage Options in AWS?

Storage options are the different services AWS provides to store data. Each service is designed for specific use cases and has different:

  • Performance
  • Availability
  • Durability
  • Replication behavior
  • Cost

2. Key Storage Services You Must Know for the Exam

2.1 Amazon S3 (Simple Storage Service)

What it is:

Object storage for files, backups, logs, media, etc.

Key Characteristics:

Durability

  • 11 9’s durability (99.999999999%)
  • Data is automatically stored across multiple devices and facilities

Availability

  • Typically 99.99% availability

Replication

  • Automatically replicates data across multiple Availability Zones (AZs) in a region
  • Optional replication:
    • Same-Region Replication (SRR)
    • Cross-Region Replication (CRR)

Storage Classes

  • Standard (frequent access)
  • Intelligent-Tiering
  • Standard-IA (Infrequent Access)
  • One Zone-IA (single AZ)
  • Glacier (archival)

2.2 Amazon EBS (Elastic Block Store)

What it is:

Block storage used with EC2 instances (like a virtual hard disk)

Key Characteristics:

Durability

  • Data is replicated within a single Availability Zone
  • Durable but not as resilient as S3 across AZs

Availability

  • Depends on the AZ
  • If AZ fails → volume becomes unavailable

Replication

  • Automatic replication inside the AZ
  • Manual replication:
    • Snapshots to S3
    • Copy snapshots across regions

Performance

  • High-performance (SSD, HDD options)

2.3 Amazon EFS (Elastic File System)

What it is:

Shared file storage for multiple EC2 instances

Key Characteristics:

Durability

  • Stored across multiple AZs automatically

Availability

  • Highly available (regional service)

Replication

  • Built-in multi-AZ replication
  • Optional cross-region replication

Use Case

  • Multiple servers accessing the same files

2.4 Amazon FSx

What it is:

Managed file systems (Windows, Lustre, NetApp ONTAP, OpenZFS)

Key Characteristics:

  • Depends on the file system type
  • Can support multi-AZ deployment
  • High performance for specialized workloads

2.5 Instance Store

What it is:

Temporary storage attached to EC2

Key Characteristics:

Durability

  • Not durable
  • Data lost when instance stops or fails

Availability

  • Only available while instance is running

Replication

  • No replication

2.6 Amazon S3 Glacier

What it is:

Low-cost archival storage

Key Characteristics:

Durability

  • Same as S3 (11 9’s)

Availability

  • Lower availability than standard S3

Retrieval

  • Slower access (minutes to hours)

3. Core Concepts You MUST Understand


3.1 Durability vs Availability

Durability

  • Measures how safe your data is from being lost
  • Example: If hardware fails, is your data still safe?

Availability

  • Measures how often your data can be accessed

Exam Tip:

  • S3 → Very high durability
  • EBS → High durability but limited to one AZ
  • Instance Store → No durability

3.2 Replication

Replication means making copies of data to improve fault tolerance.


Types of Replication in AWS

1. Within Availability Zone

  • Example: EBS replicates data inside the same AZ
  • Protects against disk failure

2. Multi-AZ Replication

  • Example:
    • S3
    • EFS
  • Protects against AZ failure

3. Cross-Region Replication

  • Example:
    • S3 CRR
    • EFS replication

Used for:

  • Disaster recovery
  • Global applications

3.3 Storage Types Comparison

FeatureS3EBSEFSInstance Store
TypeObjectBlockFileBlock
DurabilityVery HighHighVery HighLow
Multi-AZYesNoYesNo
ReplicationAuto (multi-AZ)AZ onlyMulti-AZNone
PersistenceYesYesYesNo

4. High Availability and Fault Tolerance Design

To design fault-tolerant systems, you must:


Use Multi-AZ Storage

  • S3 and EFS automatically handle AZ failures
  • EBS requires backup strategy

Add Backup and Replication

  • EBS → Use snapshots
  • S3 → Enable CRR for disaster recovery
  • Databases → Enable Multi-AZ

Avoid Single Points of Failure

  • Do NOT rely on:
    • Single EBS volume
    • Instance Store

Combine Storage with Compute

Example architecture:

  • EC2 + EBS → Use snapshots
  • Shared workloads → Use EFS
  • Static data → Use S3

5. Important Exam Scenarios


Scenario 1:

Need highest durability and multi-AZ storage

✅ Answer: S3


Scenario 2:

Need low-latency storage attached to EC2

✅ Answer: EBS


Scenario 3:

Need shared file system across multiple EC2 instances

✅ Answer: EFS


Scenario 4:

Need temporary high-speed storage

✅ Answer: Instance Store


Scenario 5:

Need disaster recovery across regions

✅ Answer:

  • S3 CRR
  • EBS snapshot copy
  • EFS replication

6. Key Exam Tips (VERY IMPORTANT)

  • S3 = Most durable storage
  • EBS = Single AZ only
  • EFS = Multi-AZ shared storage
  • Instance Store = Temporary
  • Replication improves fault tolerance
  • Snapshots = backup mechanism for EBS

7. Quick Memory Summary

  • Durability = Data safety
  • Availability = Data access
  • Replication = Copying data for protection
Buy Me a Coffee