Amazon DynamoDB Streams is a feature of DynamoDB that captures changes to items in a DynamoDB table and makes these changes available to other AWS services or applications in real time. It provides a time-ordered sequence of item-level changes (insert, modify, delete) made to a DynamoDB table.
DynamoDB Streams allows you to process data changes in real time, and can be used to trigger actions such as invoking AWS Lambda functions, replicating data to other systems, or integrating with other AWS services.
Capture Changes:
DynamoDB Streams records changes to data (such as inserts, updates, and deletes) in your DynamoDB table. Each change is captured in a stream record.
Stream Record:
Each stream record contains a sequence number, the change type (insert, modify, delete), and a timestamp for when the change occurred.
It also contains the old image (before modification) and new image (after modification) of the affected item, depending on your configuration (more on this below).
Stream Processing:
Once the changes are captured, you can process them by reading from the stream using a consumer (e.g., AWS Lambda, Kinesis, or your own application).
The consumer can perform operations based on the data change, such as triggering business logic or syncing the data with other systems.
Stream Retention:
Stream records are retained for 24 hours in the stream, and after this period, they are automatically deleted. You must consume the data within this window to process the events.