Data access patterns and consistency models

Using Data Stores in AWS

Dunieski Otano

AWS Solutions Architect

The problem with one-size-fits-all

  • Product catalog: 1000s of reads/sec, few updates
  • Shopping cart: Frequent writes
  • Order system: Cannot lose data
  • Different access patterns, require different data store strategies

users

Using Data Stores in AWS

Ephemeral vs persistent storage patterns

  • Persistent Storage: Data survives restarts

    • Stored on disk, permanent until deleted
    • AWS Examples: DynamoDB, S3, RDS, OpenSearch
    • Use for: User data, transactions, files, records

    persistent

  • Ephemeral Storage: Data lost on restart
    • Stored in memory, temporary by design
    • AWS Examples: ElastiCache, EC2 Instance Store
    • Use for: Session data, cache, temporary calculations

ephemeral

Using Data Stores in AWS

Read-heavy vs write-heavy workloads

  • Read-heavy: Many reads, few writes
    • Example: Product catalog, blog posts
    • Best for: DynamoDB, S3 with caching
  • Write-heavy: Frequent updates
    • Example: IoT sensors, clickstream data
    • Best for: DynamoDB with auto-scaling
  • Balanced: Equal read/write operations
    • Example: Social media posts, comments
    • Best for: DynamoDB with balanced capacity

bars

Using Data Stores in AWS

Consistency models in practice

  • Strongly consistent: Always returns latest data
    • Slower, but accurate
    • Use for: Financial data, inventory counts
  • Eventually consistent: May return stale data
    • Faster, but brief delays possible
    • Use for: User profiles, product descriptions
  • DynamoDB Console: ConsistentRead parameter

 

table_query

Using Data Stores in AWS

Exploring services in AWS console

  • DynamoDB Console: Tables, items, queries
  • S3 Console: Buckets, objects, properties
  • ElastiCache Console: Redis/Memcached, nodes, metrics
  • Each service organizes resources differently

services

Using Data Stores in AWS

Matching patterns to services

  • Read-heavy + eventual consistency = DynamoDB, S3
  • Write-heavy + high throughput = DynamoDB
  • Caching layer = ElastiCache
  • Full-text search = OpenSearch

consistency

Using Data Stores in AWS

Let's practice!

Using Data Stores in AWS

Preparing Video For Download...