Stream processing patterns in DynamoDB - Step-by-Step Execution

Concept Flow - Stream processing patterns

Data Change in DynamoDB Table

↓

DynamoDB Streams Capture Change

↓

Stream Processing Lambda Triggered

↓

Process Record: Filter/Transform/Enrich

↓

Write Results to Target (DB, Analytics, etc)

↓

End

Data changes in DynamoDB tables create stream records, which trigger processing functions that handle each record and write results to targets.

Execution Sample

DynamoDB

1. Insert item into DynamoDB table
2. Stream captures INSERT event
3. Lambda triggered with stream record
4. Lambda processes record (e.g., filter)
5. Lambda writes processed data to another table

Shows how a data insert triggers a stream event, which a Lambda function processes and writes results elsewhere.

Execution Table

Step	Event	Stream Record Captured	Lambda Triggered	Processing Action	Output Action
1	Insert item in table	Yes (INSERT event)	Triggered	Filter record (passes filter)	Write to target table
2	Update item in table	Yes (MODIFY event)	Triggered	Transform record (change attribute)	Write updated data
3	Delete item in table	Yes (REMOVE event)	Triggered	Log deletion	No write (optional)
4	No change	No event	No trigger	No processing	No output

💡 Execution stops when no new stream records are available or Lambda finishes processing all records.

Variable Tracker

Variable	Start	After Step 1	After Step 2	After Step 3	Final
Stream Record	Empty	INSERT event captured	MODIFY event captured	REMOVE event captured	Empty after processing
Lambda Trigger	Inactive	Active	Active	Active	Inactive
Processed Output	None	Written new item	Written updated item	Logged deletion	None

Key Moments - 3 Insights

Why does the Lambda trigger only when there is a data change?

Because DynamoDB Streams only capture changes (INSERT, MODIFY, REMOVE), so Lambda triggers only when a stream record exists, as shown in execution_table rows 1-3.

What happens if the record does not pass the filter in processing?

The Lambda skips writing output for that record, so no new data is written, which can be seen as no output action in the processing step.

Can the stream capture multiple types of events?

Yes, it captures INSERT, MODIFY, and REMOVE events, each triggering Lambda with different processing actions as shown in the execution_table.

Visual Quiz - 3 Questions

Look at the execution_table, at which step does the Lambda write updated data?

A Step 3

B Step 1

C Step 2

D Step 4

💡 Hint

Check the 'Output Action' column for each step in the execution_table.

According to variable_tracker, what is the state of 'Lambda Trigger' after Step 3?

A Inactive

B Active

C Error

D Unknown

💡 Hint

Look at the 'Lambda Trigger' row and the column 'After Step 3' in variable_tracker.

If no data change occurs, what happens to the stream and Lambda trigger?

A No stream event, Lambda does not trigger

B Stream captures event, Lambda does not trigger

C Stream captures event, Lambda triggers

D No stream event, Lambda triggers

💡 Hint

Refer to execution_table row 4 for no change scenario.

Concept Snapshot

DynamoDB Streams capture table data changes (INSERT, MODIFY, REMOVE).
Each change creates a stream record.
Lambda functions trigger on these records.
Lambda processes records (filter, transform).
Processed data can be written to other targets.
No change means no stream event and no Lambda trigger.

Full Transcript

In DynamoDB, when data changes in a table, these changes are captured as stream records. These records represent events like inserts, updates, or deletes. A Lambda function can be set to trigger automatically when these stream records appear. The Lambda processes each record, for example by filtering or transforming the data, and then writes the results to another database or system. If there is no data change, no stream record is created, so the Lambda does not trigger. This flow ensures that processing happens only when data changes, making it efficient and event-driven.