RDS vs DynamoDB

RDS vs DynamoDB

Step 1 of 10

tableorders

Every DynamoDB item has a partition key. Three customers are about to write orders.

The Idea

RDS runs managed relational engines — MySQL, PostgreSQL, MariaDB, Oracle, SQL Server, Db2 — for joins, transactions and ad-hoc SQL. Multi-AZ keeps a synchronous standby for failover; asynchronous read replicas scale reads. DynamoDB is key-value: the partition key is hashed to choose a partition, and an optional sort key orders items under one key.

Real-World Example

An orders table keyed by customer ID, sorted by order date, makes "this customer's orders, newest first" one Query. Key every flash-sale write by the sale ID instead, and all traffic lands on one hot partition; a key suffix (write sharding) spreads it.

The Tradeoff

Each partition tops out at 3,000 read units and 1,000 write units per second (as of September 2026), so a hot key throttles while the table has room. DynamoDB wants access patterns known up front; RDS forgives a query nobody planned.

Hands-On

# illustrative — needs AWS credentials and an "orders" table
# (partition key customer_id, sort key order_ts)
import boto3
from boto3.dynamodb.conditions import Key

table = boto3.resource("dynamodb").Table("orders")
table.put_item(Item={"customer_id": "c#42", "order_ts": "2026-09-28T10:00", "total": 30})
resp = table.query(
    KeyConditionExpression=Key("customer_id").eq("c#42"),
    ScanIndexForward=False,  # newest first
)

Your turn

Put the steps in the right order.

1. 1Append a suffix 1..N to the key so writes hash to different partitions↑↓
2. 2Notice every write in the sale uses one partition key value↑↓
3. 3Read the sale back by querying each suffix and merging↑↓
4. 4See one partition throttle while the table has spare capacity↑↓

Check order

Mini quiz

1 / 3

DynamoDB decides which partition stores an item by:

• AThe sort key
• BHashing the partition key
• CInsertion order
• DThe item's size

Sources

• Core components of Amazon DynamoDB — DynamoDB Developer Guide
• Best practices for designing partition keys — DynamoDB Developer Guide
• Using write sharding to distribute workloads evenly — DynamoDB Developer Guide
• Multi-AZ DB instance deployments — Amazon RDS User Guide