Why single-table design matters in DynamoDB - Performance Analysis

Time Complexity: Why single-table design matters

O(n)

Understanding Time Complexity

When using DynamoDB, how we organize data affects how fast queries run.

We want to see how single-table design changes the work DynamoDB does as data grows.

Scenario Under Consideration

Analyze the time complexity of querying data in a single-table design.

    {
      TableName: "MyTable",
      KeyConditionExpression: "PK = :pkValue",
      ExpressionAttributeValues: {
        ":pkValue": "USER#123"
      }
    }
    

This query fetches all items with a partition key for one user in a single table.

Identify Repeating Operations

Look at what repeats when DynamoDB runs this query.

• Primary operation: Reading items with the same partition key.
• How many times: Once per item matching the key, sequentially.

How Execution Grows With Input

As the number of items for a user grows, the work grows too.

Input Size (n)	Approx. Operations
10	10 reads
100	100 reads
1000	1000 reads

Pattern observation: The work grows directly with the number of items for that key.

Final Time Complexity

Time Complexity: O(n)

This means the time to get all items grows in a straight line with how many items match the key.

Common Mistake

[X] Wrong: "Using a single table means queries are always faster no matter what."

[OK] Correct: Even with one table, reading many items for a key takes more time because each item must be read.

Interview Connect

Understanding how data layout affects query speed shows you think about real database work and efficiency.

Self-Check

What if we added a sort key to narrow queries? How would that change the time complexity?

Practice

1. Why is single-table design important in DynamoDB?

easy

A. It requires more tables to organize data.

B. It stores related data together for faster access.

C. It increases the number of database calls.

D. It makes key design less important.

Solution

1. Step 1: Understand single-table design purpose

   Single-table design groups related data in one table to reduce complexity and speed up queries.

2. Step 2: Compare options with this purpose

   Only It stores related data together for faster access. correctly states it stores related data together for faster access.

3. Final Answer:

   It stores related data together for faster access. -> Option B

4. Quick Check:

   Single-table design = faster access [OK]

Hint: Single-table means fewer tables, faster queries [OK]

Common Mistakes:

• Thinking it increases database calls
• Believing it needs more tables
• Ignoring key design importance

2. Which of the following is the correct way to define a partition key in DynamoDB single-table design?

easy

A. PartitionKey: userId

B. PartitionKey: {userId, orderId}

C. PartitionKey: userId + orderId

D. PartitionKey: orderId

Solution

1. Step 1: Recall partition key definition

   The partition key must be a single attribute that uniquely partitions data.

2. Step 2: Analyze options

   PartitionKey: userId uses a single attribute userId, which is valid. PartitionKey: {userId, orderId} tries to use multiple attributes as one key, which is invalid syntax. PartitionKey: userId + orderId concatenates keys but is not a direct attribute definition. PartitionKey: orderId uses orderId alone, which may not be unique.

3. Final Answer:

   <code>PartitionKey: userId</code> -> Option A

4. Quick Check:

   Partition key = single attribute [OK]

Hint: Partition key must be a single attribute [OK]

Common Mistakes:

• Using multiple attributes directly as partition key
• Concatenating keys without defining attribute
• Choosing non-unique keys

3. Given a single-table design with partition key UserId and sort key OrderId, what will this query return?

SELECT * FROM Orders WHERE UserId = '123' AND OrderId = 'A1'

medium

A. All orders for user 123

B. All orders with OrderId 'A1' regardless of user

C. Only the order with OrderId 'A1' for user 123

D. Syntax error in query

Solution

1. Step 1: Understand key conditions in query

   The query filters by partition key UserId = '123' and sort key OrderId = 'A1', so it targets one specific item.

2. Step 2: Determine result based on keys

   Since both keys are specified, only the single matching order for user 123 with OrderId 'A1' will be returned.

3. Final Answer:

   Only the order with OrderId 'A1' for user 123 -> Option C

4. Quick Check:

   Partition + sort key = single item [OK]

Hint: Specify both keys to get one item [OK]

Common Mistakes:

• Assuming query returns all orders for user
• Ignoring sort key in query
• Thinking query syntax is invalid

4. You designed a single-table with partition key UserId and sort key OrderId. Your query to get all orders for a user returns no results, but you know data exists. What is the likely mistake?

medium

A. You queried only by sort key without partition key.

B. You used the wrong table name in the query.

C. You queried only by partition key without sort key.

D. You forgot to create the table.

Solution

1. Step 1: Recall query requirements for partition and sort keys

   In DynamoDB, queries must specify the partition key. Querying only by sort key is invalid and returns no results.

2. Step 2: Analyze options

   You queried only by sort key without partition key. matches the mistake of querying only by sort key. You used the wrong table name in the query. and D are possible but less likely if data exists. You queried only by partition key without sort key. is valid and returns all items with that partition key.

3. Final Answer:

   You queried only by sort key without partition key. -> Option A

4. Quick Check:

   Query must include partition key [OK]

Hint: Always include partition key in queries [OK]

Common Mistakes:

• Querying only by sort key
• Using wrong table name
• Assuming table not created

5. In a single-table design, you want to store both user profiles and their orders in one table. Which key design helps you efficiently query all orders for a user and also get the user profile quickly?

hard

A. Partition key: UserId#OrderId, Sort key: ItemType

B. Partition key: OrderId, Sort key: UserId

C. Partition key: ItemType, Sort key: UserId#OrderId

D. Partition key: UserId, Sort key: ItemType#OrderId

Solution

1. Step 1: Understand single-table design key patterns

   Using a partition key as UserId groups all user data together. Adding a sort key with ItemType#OrderId lets you separate user profile and orders.

2. Step 2: Evaluate options for efficient queries

   Partition key: UserId, Sort key: ItemType#OrderId allows querying all orders by filtering sort key starting with 'Order' and fetching profile with sort key 'Profile'. Other options mix keys making queries inefficient or impossible.

3. Final Answer:

   Partition key: UserId, Sort key: ItemType#OrderId -> Option D

4. Quick Check:

   Partition by user, sort by item type [OK]

Hint: Use partition key for user, sort key to separate items [OK]

Common Mistakes:

• Using orderId as partition key
• Not distinguishing item types in sort key
• Combining keys incorrectly