DynamoDB Data Modeling: Partition and Sort Keys

Indexing and Queries

DynamoDB provides indexing options to optimize query flexibility:

Global Secondary Index (GSI)

• Allows querying by different keys from the base table;
• Ideal when you need multiple ways to access data (e.g., by username instead of user ID).

Local Secondary Index (LSI)

• Uses the same partition key, but a different sort key;
• Useful for filtering data within a single partition based on other attributes.

Advanced Features

Transactions

• Enables atomic operations across multiple items and tables;
• Ensures data consistency in financial, inventory, or user-related systems.

Time to Live (TTL)

• Attach an expiry timestamp to records;
• DynamoDB will automatically delete expired items — perfect for session tokens, temporary cache, or logs.

DynamoDB Streams

DynamoDB Streams captures every item-level change (insert, update, delete) in near real-time and stores them for 24 hours.

You can:

• Trigger AWS Lambda functions in response;
• Build event-driven workflows like sending notifications, syncing data, updating analytics dashboards, or even audit trails;
• Avoid manual polling, reducing overhead.

Creating a Table from the AWS CLI

Let's walk through creating a table called orders using the AWS CLI:

aws dynamodb create-table \
  --table-name orders \
  --attribute-definitions \
    AttributeName=user_id,AttributeType=S \
    AttributeName=order_id,AttributeType=S \
  --key-schema \
    AttributeName=user_id,KeyType=HASH \
    AttributeName=order_id,KeyType=RANGE \
  --provisioned-throughput \
    ReadCapacityUnits=5 \
    WriteCapacityUnits=5

• user_id: partition key;
• order_id: sort key;
• Provisioned Throughput: setting read and write capacity manually (more on this below).

Once created, verify it:

aws dynamodb describe-table --table-name orders

And finally, to clean up and avoid charges:

aws dynamodb delete-table --table-name orders

Capacity Modes

DynamoDB supports two capacity modes:

Provisioned Mode

• You manually specify read and write units;
• Suitable for predictable workloads;
• Supports auto-scaling to adjust capacity based on usage.

On-Demand Mode

• No upfront capacity planning;
• Pay per request (read/write);
• Ideal for spiky or unpredictable traffic.

Monitoring and Optimization

Use Amazon CloudWatch to:

• Track performance metrics;
• Identify throttling events, which signal your application is exceeding the provisioned throughput;
• Optimize access patterns or scale up as needed.

Summary

Amazon DynamoDB is:

• Fast: consistently low-latency reads and writes;
• Scalable: handles millions of operations seamlessly;
• Serverless: zero infrastructure to manage;
• Feature-rich: with streams, TTL, indexes, and transactions.