Scan vs Query performance comparison in DynamoDB - Practice Questions

Challenge - 5 Problems

🧠 Conceptual

intermediate

Understanding Scan vs Query in DynamoDB

Which statement best describes the main difference between a Scan and a Query operation in DynamoDB?

A Scan reads all items in a table and filters results, while Query finds items based on primary key values.

B Scan only reads items with a specific attribute, Query reads all items.

C Scan and Query both read all items but Query is faster because it uses indexes.

D Query reads all items in a table and filters results, while Scan finds items based on primary key values.

💡 Hint

Think about how each operation accesses data and what keys they use.

A Scan reads every item in the table and then applies filters, which can be slow. A Query uses the primary key or an index to find matching items quickly.

❓ query_result

intermediate

Result count difference between Scan and Query

Given a DynamoDB table with 1000 items, 100 of which have a partition key value of 'user123'. What is the expected number of items returned by a Query with partition key 'user123' vs a Scan with a filter for 'user123'?

A Query returns 100 items; Scan returns 1000 items.

B Query returns 1000 items; Scan returns 100 items.

C Query returns 100 items; Scan reads 1000 items but returns 100 after filtering.

D Query returns 1000 items; Scan reads 1000 items but returns 100 after filtering.

💡 Hint

Consider how many items each operation reads and returns.

The Query directly fetches the 100 matching items. The Scan reads all 1000 items and then filters to return 100.

📝 Syntax

advanced

Identify the correct Query syntax in DynamoDB

Which of the following DynamoDB Query API calls is syntactically correct to fetch items with partition key 'user123'?

DynamoDB

const params = {
  TableName: 'Users',
  KeyConditionExpression: 'userId = :uid',
  ExpressionAttributeValues: {
    ':uid': 'user123'
  }
};

const result = await dynamodb.query(params).promise();

A FilterExpression: 'userId = :uid', ExpressionAttributeValues: { ':uid': 'user123' }

B KeyConditionExpression: 'userId == :uid', ExpressionAttributeValues: { ':uid': 'user123' }

C KeyConditionExpression: 'userId = user123', ExpressionAttributeValues: { ':uid': 'user123' }

D KeyConditionExpression: 'userId = :uid', ExpressionAttributeValues: { ':uid': 'user123' }

💡 Hint

Check the syntax for KeyConditionExpression and how values are referenced.

KeyConditionExpression uses a single equals sign and placeholders like ':uid' must be defined in ExpressionAttributeValues. Option D follows this correctly.

❓ optimization

advanced

Improving Scan performance with filters

You want to reduce the read capacity units consumed by a Scan operation that filters items by an attribute 'status'. Which approach will improve performance?

A Use a FilterExpression in Scan to filter 'status' and enable Parallel Scan with segments.

B Use Query instead of Scan without any indexes.

C Remove the FilterExpression and scan the entire table sequentially.

D Use Scan with FilterExpression and set ConsistentRead to true.

💡 Hint

Think about how to reduce the amount of data scanned and speed up the operation.

FilterExpression reduces returned items but does not reduce read units. Parallel Scan divides the table into segments scanned concurrently, improving speed.

🔧 Debug

expert

Diagnosing slow Query performance

A DynamoDB Query on a large table is unexpectedly slow. The Query uses a partition key but no sort key condition. Which is the most likely cause?

A The Query uses a FilterExpression instead of KeyConditionExpression, causing a syntax error.

B The Query returns too many items because no sort key condition limits results, causing high latency.

C The table has no provisioned throughput, so Query is throttled.

D The partition key is missing from the KeyConditionExpression, causing a full table scan.

💡 Hint

Consider how the absence of a sort key condition affects the number of items returned.

Without a sort key condition, the Query returns all items with the partition key, which can be many and slow the response.