BatchGetItem in DynamoDB - Step-by-Step Execution

Concept Flow - BatchGetItem

Start BatchGetItem Request

↓

Receive list of keys

↓

Check keys per table

↓

Send batch request to DynamoDB

↓

DynamoDB returns items and unprocessed keys

↓

If unprocessed keys exist?

Yes→Retry unprocessed keys

↓

Repeat until no unprocessed keys

↓

Return all retrieved items

↓

End

BatchGetItem sends multiple keys to DynamoDB, retrieves items in batches, retries unprocessed keys, and returns all found items.

Execution Sample

DynamoDB

BatchGetItem({
  RequestItems: {
    'Table1': { Keys: [{ id: 1 }, { id: 2 }] },
    'Table2': { Keys: [{ id: 3 }] }
  }
})

This request asks DynamoDB to get items with id 1 and 2 from Table1 and id 3 from Table2 in one batch.

Execution Table

Step	Action	Keys Sent	Response	Unprocessed Keys	Notes
1	Send batch request	Table1: [1,2], Table2: [3]	Items for ids 1,2,3 returned	None	All keys processed in first try
2	Return items	N/A	Items for ids 1,2,3	N/A	BatchGetItem completes successfully

💡 All requested keys processed, no unprocessed keys remain

Variable Tracker

Variable	Start	After Step 1	After Step 2
Keys to get	[Table1:1,2; Table2:3]	[None]	[None]
Unprocessed Keys	[All keys]	[None]	[None]
Retrieved Items	[]	[Items for 1,2,3]	[Items for 1,2,3]

Key Moments - 2 Insights

Why do we sometimes get unprocessed keys back from DynamoDB?

DynamoDB limits batch size and throughput. If it can't process all keys at once, it returns unprocessed keys to retry later, as shown in the 'Unprocessed Keys' column in the execution_table.

What happens if there are unprocessed keys after the first request?

The client retries only those unprocessed keys until all are processed, ensuring no data is missed. This retry loop is part of the flow in the concept_flow diagram.

Visual Quiz - 3 Questions

Look at the execution_table, what keys are sent in the first batch request?

A Table1: [1], Table2: [3]

B Table1: [2], Table2: []

C Table1: [1,2], Table2: [3]

D Table1: [], Table2: [3]

💡 Hint

Check the 'Keys Sent' column in Step 1 of the execution_table

At which step does BatchGetItem return the retrieved items?

A Step 1

B Step 2

C Before Step 1

D After Step 3

💡 Hint

Look at the 'Action' and 'Response' columns in the execution_table

If DynamoDB returns unprocessed keys, what should the client do next?

A Retry only the unprocessed keys

B Retry all keys again

C Ignore them and finish

D Stop and report error

💡 Hint

Refer to the concept_flow where unprocessed keys lead to retry

Concept Snapshot

BatchGetItem lets you get multiple items from one or more tables in a single request.
You send a list of keys per table.
DynamoDB returns found items and may return unprocessed keys.
You retry unprocessed keys until all are retrieved.
This reduces multiple calls and improves efficiency.

Full Transcript

BatchGetItem is a DynamoDB operation to fetch multiple items by their keys in one request. You provide keys grouped by table. DynamoDB tries to return all items at once but may return some keys as unprocessed if limits are hit. The client then retries those keys until all are processed. This process reduces network calls and speeds up data retrieval. The execution flow starts with sending keys, receiving items and unprocessed keys, retrying if needed, and finally returning all items. Variables like keys to get, unprocessed keys, and retrieved items change during execution. Understanding retries and unprocessed keys is key to using BatchGetItem effectively.

Practice

1. What is the main purpose of the BatchGetItem operation in DynamoDB?

easy

A. To update multiple items in a table at once.

B. To retrieve multiple items from one or more tables using their primary keys in a single request.

C. To delete multiple items from a table in one call.

D. To scan the entire table and return all items.

Solution

1. Step 1: Understand BatchGetItem functionality

   BatchGetItem is designed to fetch multiple items quickly by specifying their primary keys.

2. Step 2: Compare with other operations

   Unlike update or delete, BatchGetItem only retrieves data, not modifies it.

3. Final Answer:

   To retrieve multiple items from one or more tables using their primary keys in a single request. -> Option B

4. Quick Check:

   BatchGetItem = fetch multiple items [OK]

Hint: BatchGetItem fetches multiple items by keys in one call [OK]

Common Mistakes:

• Confusing BatchGetItem with update or delete operations
• Thinking it scans the whole table
• Assuming it modifies data

2. Which of the following is the correct syntax snippet to specify keys in a BatchGetItem request in DynamoDB?

easy

A. "RequestItems": { "TableName": { "FilterExpression": "Id = :id" } }

B. "RequestItems": { "TableName": { "AttributesToGet": ["Id"] } }

C. "RequestItems": { "TableName": { "Keys": [{ "Id": { "S": "123" } }] } }

D. "RequestItems": { "TableName": { "UpdateExpression": "SET Name = :name" } }

Solution

1. Step 1: Identify BatchGetItem key specification

   BatchGetItem requires a "Keys" array with primary key objects for each item to fetch.

2. Step 2: Check each option

   "RequestItems": { "TableName": { "Keys": [{ "Id": { "S": "123" } }] } } correctly uses "Keys" with primary key format. Snippets using "AttributesToGet", "FilterExpression", or "UpdateExpression" are not valid for BatchGetItem keys.

3. Final Answer:

   "RequestItems": { "TableName": { "Keys": [{ "Id": { "S": "123" } }] } } -> Option C

4. Quick Check:

   Keys array = correct syntax [OK]

Hint: BatchGetItem keys go inside "Keys" array per table [OK]

Common Mistakes:

• Using FilterExpression instead of Keys
• Confusing AttributesToGet with Keys
• Trying to use UpdateExpression in BatchGetItem

3. Given the following BatchGetItem request, what will be the output if the table contains items with Ids 101 and 102 only?

{
  "RequestItems": {
    "Products": {
      "Keys": [
        { "Id": { "N": "101" } },
        { "Id": { "N": "103" } }
      ]
    }
  }
}

medium

A. Returns only item with Id 101

B. Returns all items in the Products table

C. Returns no items because 103 does not exist

D. Returns items with Id 101 and 103

Solution

1. Step 1: Understand BatchGetItem behavior with missing keys

   BatchGetItem returns only items that exist for the specified keys. Missing keys are ignored.

2. Step 2: Analyze given keys

   Id 101 exists, so it will be returned. Id 103 does not exist, so it will be skipped.

3. Final Answer:

   Returns only item with Id 101 -> Option A

4. Quick Check:

   Only existing keys returned [OK]

Hint: BatchGetItem skips missing keys, returns existing only [OK]

Common Mistakes:

• Expecting error if key not found
• Assuming all keys must exist
• Thinking it returns all table items

4. You wrote this BatchGetItem request but it returns a syntax error:

{
  "RequestItems": {
    "Orders": {
      "Keys": {
        "OrderId": { "S": "A123" }
      }
    }
  }
}

What is the error and how to fix it?

medium

A. RequestItems key name is misspelled.

B. OrderId should be a number, not a string.

C. BatchGetItem does not support multiple tables.

D. Keys must be an array of key objects, not a single object.

Solution

1. Step 1: Identify the Keys format error

   Keys must be an array of objects, but here it is a single object, causing syntax error.

2. Step 2: Correct the Keys format

   Wrap the key object inside an array: "Keys": [{ "OrderId": { "S": "A123" } }]

3. Final Answer:

   Keys must be an array of key objects, not a single object. -> Option D

4. Quick Check:

   Keys = array of key objects [OK]

Hint: Keys must always be an array, even for one item [OK]

Common Mistakes:

• Using object instead of array for Keys
• Wrong data type for key attribute
• Misspelling RequestItems

5. You want to fetch items from two tables, "Users" and "Orders", using BatchGetItem. You have 3 user IDs and 2 order IDs. Which of the following request structures correctly fetches all these items in one call?

hard

A. { "RequestItems": { "Users": { "Keys": [ { "UserId": { "S": "U1" } }, { "UserId": { "S": "U2" } }, { "UserId": { "S": "U3" } } ] }, "Orders": { "Keys": [ { "OrderId": { "S": "O1" } }, { "OrderId": { "S": "O2" } } ] } } }

B. { "RequestItems": { "Users": { "Keys": { "UserId": ["U1", "U2", "U3"] } }, "Orders": { "Keys": { "OrderId": ["O1", "O2"] } } } }

C. { "RequestItems": { "Users": { "FilterExpression": "UserId IN (:u1, :u2, :u3)" }, "Orders": { "FilterExpression": "OrderId IN (:o1, :o2)" } } }

D. { "RequestItems": { "Users": { "AttributesToGet": ["UserId"] }, "Orders": { "AttributesToGet": ["OrderId"] } } }

Solution

1. Step 1: Understand BatchGetItem multi-table request format

   BatchGetItem requires "RequestItems" with each table name mapping to a "Keys" array of key objects.

2. Step 2: Evaluate each option

   { "RequestItems": { "Users": { "Keys": [ { "UserId": { "S": "U1" } }, { "UserId": { "S": "U2" } }, { "UserId": { "S": "U3" } } ] }, "Orders": { "Keys": [ { "OrderId": { "S": "O1" } }, { "OrderId": { "S": "O2" } } ] } } } correctly uses arrays of key objects for both tables. Formats using arrays inside objects, FilterExpression, or AttributesToGet are invalid for BatchGetItem keys.

3. Final Answer:

   { "RequestItems": { "Users": { "Keys": [ { "UserId": { "S": "U1" } }, { "UserId": { "S": "U2" } }, { "UserId": { "S": "U3" } } ] }, "Orders": { "Keys": [ { "OrderId": { "S": "O1" } }, { "OrderId": { "S": "O2" } } ] } } } -> Option A

4. Quick Check:

   Keys array per table = correct multi-table BatchGetItem [OK]

Hint: Use Keys array per table for multi-table BatchGetItem [OK]

Common Mistakes:

• Using FilterExpression instead of Keys
• Passing keys as arrays inside objects incorrectly
• Using AttributesToGet instead of Keys