BatchWriteItem in DynamoDB - Step-by-Step Execution

Concept Flow - BatchWriteItem

Prepare batch request with multiple Put/Delete

↓

Send BatchWriteItem request to DynamoDB

↓

DynamoDB processes items

↓

Check for UnprocessedItems

↓

Retry UnprocessedItems

BatchWriteItem sends multiple write requests at once. DynamoDB processes them and may return unprocessed items to retry.

Execution Sample

DynamoDB

BatchWriteItem({
  RequestItems: {
    'Table1': [
      { PutRequest: { Item: { id: '1', name: 'A' } } },
      { DeleteRequest: { Key: { id: '2' } } }
    ]
  }
})

This batch request writes one new item and deletes another in Table1.

Execution Table

Step	Action	Items Sent	DynamoDB Response	Next Step
1	Send BatchWriteItem request	Put id=1, Delete id=2	Processed both items	No UnprocessedItems, finish
2	Finish	-	-	Batch write complete

💡 All items processed successfully, no retries needed

Variable Tracker

Variable	Start	After Step 1	Final
RequestItems	{Put id=1, Delete id=2}	Sent to DynamoDB	No UnprocessedItems
UnprocessedItems	{}	{}	{}

Key Moments - 2 Insights

Why do we check for UnprocessedItems after sending the batch?

Because DynamoDB might not process all items in one go due to limits. The execution_table row 1 shows checking UnprocessedItems to decide if retry is needed.

Can BatchWriteItem mix Put and Delete requests?

Yes, as shown in the execution_sample code and execution_table row 1, both PutRequest and DeleteRequest can be in the same batch.

Visual Quiz - 3 Questions

Look at the execution_table, what happens if there are UnprocessedItems after the first request?

A The operation fails immediately

B The entire batch is resent

C The client retries sending only the unprocessed items

D DynamoDB processes unprocessed items automatically

💡 Hint

Refer to concept_flow where UnprocessedItems lead to retry of only those items

According to variable_tracker, what is the state of UnprocessedItems after step 1 in this example?

A Empty object, no retries needed

B Contains items to retry

C Undefined

D Contains all original items

💡 Hint

Check variable_tracker row for UnprocessedItems after Step 1

If we add more items than DynamoDB can process at once, what will the execution_table show?

A UnprocessedItems will be empty

B Some items will appear in UnprocessedItems to retry

C DynamoDB will process all items without issue

D BatchWriteItem will reject the request

💡 Hint

Recall concept_flow and key_moments about retrying unprocessed items

Concept Snapshot

BatchWriteItem lets you write or delete multiple items in one request.
You send a batch of PutRequest and DeleteRequest objects.
DynamoDB may return UnprocessedItems if it can't handle all at once.
You must retry only those unprocessed items.
This improves efficiency over single writes.

Full Transcript

BatchWriteItem is a DynamoDB operation to write or delete multiple items in one request. You prepare a batch with PutRequest and DeleteRequest objects for your table. When you send this batch, DynamoDB tries to process all items. Sometimes, due to limits, some items are not processed and returned as UnprocessedItems. You check for these and retry only those items. This process repeats until all items are processed or you decide to stop. This method is efficient because it reduces the number of requests compared to writing items one by one.