Hierarchical data modeling in DynamoDB - Practice Problems & Coding Challenges

Challenge - 5 Problems

❓ query_result

intermediate

Query to retrieve all child items of a parent

You have a DynamoDB table storing categories and subcategories. Each item has a PK as the parent category ID and a SK as the child category ID.

Which query will return all subcategories of the parent category with PK = 'CAT#100'?

DynamoDB

TableName: Categories
PartitionKey: PK
SortKey: SK
Example item: { PK: 'CAT#100', SK: 'SUB#200', Name: 'Subcategory A' }

A Query with KeyConditionExpression: "PK = :pkval" and ExpressionAttributeValues: {":pkval": "CAT#100"}

B Scan with FilterExpression: "PK = :pkval" and ExpressionAttributeValues: {":pkval": "CAT#100"}

C Scan with FilterExpression: "SK = :skval" and ExpressionAttributeValues: {":skval": "SUB#200"}

D Query with KeyConditionExpression: "SK = :skval" and ExpressionAttributeValues: {":skval": "SUB#200"}

💡 Hint

Remember, Query operations use the partition key to efficiently find items.

Querying by the partition key PK returns all items with that parent ID. Scans are less efficient and not needed here. Querying by SK alone is invalid because SK is the sort key, not the partition key.

🧠 Conceptual

intermediate

Best way to model parent-child relationships in DynamoDB

Which DynamoDB design pattern is best suited to model hierarchical data with multiple levels of parent-child relationships?

A Use a single table with only a partition key and no sort key

B Use multiple tables, one per hierarchy level, linked by foreign keys

C Use a single table with composite keys where <code>PK</code> is the root ancestor and <code>SK</code> encodes the path to the child

D Use a global secondary index on a timestamp attribute

💡 Hint

Think about how DynamoDB's single-table design supports hierarchical queries.

Using composite keys with PK as the root ancestor and SK encoding the path allows efficient queries for all descendants. Multiple tables or missing sort keys do not support hierarchical queries well. GSI on timestamp is unrelated.

📝 Syntax

advanced

Identify the syntax error in this DynamoDB query expression

What is wrong with this DynamoDB query KeyConditionExpression?

KeyConditionExpression: "PK = :pkval AND begins_with(SK, :skprefix)"
ExpressionAttributeValues: {":pkval": "CAT#100", ":skprefix": "SUB#"}

A PK must be compared with SK in KeyConditionExpression

B begins_with function requires two attribute names, not attribute value

C ExpressionAttributeValues keys must not start with a colon

D The AND operator cannot be used in KeyConditionExpression

💡 Hint

Check the rules for KeyConditionExpression syntax in DynamoDB.

In DynamoDB, KeyConditionExpression supports equality condition on the partition key optionally connected with lowercase 'and' to one condition on the sort key using operators like begins_with. Uppercase 'AND' is invalid syntax.

❓ optimization

advanced

Optimize hierarchical queries with large datasets

You have a DynamoDB table with millions of hierarchical items. Queries for all descendants of a parent are slow. Which optimization will improve query performance?

A Use Scan operation with FilterExpression to find descendants

B Use a composite sort key encoding the full path and query with begins_with on SK

C Add a global secondary index on the child item name attribute

D Store all descendants in a single item as a JSON array

💡 Hint

Think about how sort keys and begins_with can limit query scope.

Encoding the full path in the sort key and querying with begins_with limits the query to relevant items efficiently. GSI on name won't help hierarchical queries. Scan is slow for large data. Storing all descendants in one item can cause size limits and slow reads.

🔧 Debug

expert

Why does this hierarchical query return no results?

Given this DynamoDB query:

KeyConditionExpression: "PK = :pkval and SK > :skval"
ExpressionAttributeValues: {":pkval": "CAT#100", ":skval": "SUB#100"}

But no items are returned, even though items with SK values greater than "SUB#100" exist. What is the most likely cause?

A The SK values are strings but the comparison is lexicographical and does not match expected order

B SK attribute is stored as a number, but compared as a string

C The comparison operator > is not supported in KeyConditionExpression

D The partition key value "CAT#100" does not exist in the table

💡 Hint

Consider how string comparisons work in DynamoDB sort keys.

DynamoDB compares sort keys lexicographically (alphabetical order). If SK values are strings like "SUB#100", "SUB#2", lex order means "SUB#2" > "SUB#100" is false because '1' < '2' in string comparison. This causes unexpected query results.