DynamoDB Streams

 

DynamoDB Streams are one of the most powerful—but often under‑explained—features in DynamoDB.

A DynamoDB Stream captures every change to items in a table:

• INSERT
• MODIFY
• REMOVE

You can then process these changes using:

• AWS Lambda
• Kinesis
• Custom consumers

This is how you build event‑driven architectures with DynamoDB.

Enable Streams on the Table:

You choose a stream view type, e.g.:

• NEW_IMAGE → only the new item
• OLD_IMAGE → only the old item
• NEW_AND_OLD_IMAGES → both
• KEYS_ONLY → only keys

I have table created by the name called "order".

I have enabled "Streams" to capture the changes to the table.

And I have added the lambda function to process the streams.

Lambda gets triggered, whenever there is a change in the DB items.

Below is the lambda snippet:

import json

def lambda_handler(event, context):

    for record in event["Records"]:

        event_name = record["eventName"]

        keys = record["dynamodb"]["Keys"]

        if event_name == "INSERT":

            new = record["dynamodb"]["NewImage"]

            print("New order:", new)

        elif event_name == "MODIFY":

            old = record["dynamodb"]["OldImage"]

            new = record["dynamodb"]["NewImage"]

            print("Order updated:", old, "→", new)

        elif event_name == "REMOVE":

            old = record["dynamodb"]["OldImage"]

            print("Order deleted:", old)

It captures the DynamoDB changes like "INSERT", "MODIFY" and "REMOVE".

Let's modify the order_name from Sam to John.

Once we made the changes, we can look at the CloudWatch logs for the stream changes.