Data Store Types and Concepts in AWS

Key Considerations

• ACID provides strict consistency but may face performance challenges at scale.

• BASE isn't inconsistent, but it's just differently consistent; it prioritizes availability and parallel processing.

Types of Data Stores

1. Persistent Data Store

• Data is durable and persists across reboots or power cycles.

• Examples:

  • Amazon Glacier.

  • Amazon RDS (Relational Database Service).

2. Transient Data Store

• Data is temporarily stored and passed through to another process.

• Examples:

  • Amazon SQS (Simple Queue Service)

  • Amazon SNS (Simple Notification Service)

  • Amazon Event Bridge (Event Routing and serveless event bus)

3. Ephemeral Data Store

• Data is temporary and does not persist when instances are stopped or services rebooted.

• Examples:

  • EC2 Instance Store.

  • Memcached (under Elastic Cache portfolio).

Performance Metrics

IOPS (Input/Output Operations Per Second)

• Measures how fast we can read and write to a device.

• Analogous to a small, fast sports car:

  • High speed.

  • Smaller data payload.

  • Optimized for quick operations.

Throughput

• Measures how much data can be moved at once.

• Analogous to a dump truck:

  • Larger data payload.

  • Slower speed.

  • Optimized for bulk data movement.

Consistency Models

ACID Model

Primarily used in relational databases, implementing features like row locking and rollbacks.

Characteristics:

• Atomic: Transactions are all-or-nothing.

• Consistent: Transactions from one valid state to another, and uncorrupted.

• Isolated: Transactions don't interfere with each other.

• Durable: Committed transactions persist reliably.

BASE Model

Alternative approach focused on scalability.

Characteristics:

• Basically Available: Prioritizes availability, even with stale data.

• Soft State: May not be instantly consistent across data stores.

• Eventually Consistent: All data stores will eventually sync and update.

Why Choose BASE over ACID?

• ACID models face scalability challenges.

• BASE accommodates parallel operations more effectively.

• Useful for services requiring high scalability.

• Examples of BASE implementation:

  • Amazon S3.

  • Amazon DynamoDB.