Schema design for read-heavy workloads in MongoDB - Practice Problems & Coding Challenges

Challenge - 5 Problems

🧠 Conceptual

intermediate

Why embed documents in MongoDB for read-heavy workloads?

In MongoDB, when designing a schema for a read-heavy application, why is embedding related data inside a single document often preferred over referencing separate documents?

A Embedding reduces the number of queries needed to fetch related data, improving read performance.

B Embedding increases the size of each document, which slows down reads significantly.

C Embedding forces MongoDB to create multiple indexes, which hurts read speed.

D Embedding requires more disk space, so it is avoided in read-heavy workloads.

💡 Hint

Think about how many database calls are needed to get all related data.

Embedding related data in one document means MongoDB can retrieve all needed information in a single read operation, which is faster for read-heavy workloads.

❓ query_result

intermediate

Result of a MongoDB query with embedded documents

Given a collection users where each user document embeds an array of orders, what will be the output of this query?

db.users.find({"orders.status": "shipped"}, {name: 1, orders: 1})

Assuming the collection has:

{ "name": "Alice", "orders": [{"id": 1, "status": "shipped"}, {"id": 2, "status": "pending"}] }
{ "name": "Bob", "orders": [{"id": 3, "status": "pending"}] }

A [{ "name": "Alice", "orders": [{"id": 1, "status": "shipped"}] }]

B []

C [{ "name": "Bob", "orders": [{"id": 3, "status": "pending"}] }]

D [{ "name": "Alice", "orders": [{"id": 1, "status": "shipped"}, {"id": 2, "status": "pending"}] }]

💡 Hint

Remember that MongoDB returns the whole document if any embedded document matches the query.

The query matches users with any order having status 'shipped'. It returns the full user document with all orders, not just the matching orders.

📝 Syntax

advanced

Correct schema design for fast reads with large arrays

Which MongoDB schema design is best for a read-heavy workload where each user has thousands of comments, and you want to quickly fetch the latest 10 comments?

A Embed all comments inside the user document in a single array.

B Store comments in a separate collection with a user_id field and index on user_id and timestamp.

C Embed comments in multiple arrays inside the user document, splitting by year.

D Store comments in a separate collection without any indexes.

💡 Hint

Think about document size limits and efficient querying.

Embedding thousands of comments in one document risks exceeding size limits and slows reads. Storing comments separately with proper indexes allows fast queries for latest comments.

❓ optimization

advanced

Optimizing read performance with indexes in MongoDB

You have a collection products with fields category, price, and rating. Your app frequently queries products by category and sorts by price ascending. Which index will optimize these queries best?

A Create a compound index on { price: 1, category: 1 }

B Create a single index on price only

C Create a compound index on { category: 1, price: 1 }

D Create a single index on category only

💡 Hint

Consider the order of fields in compound indexes for filtering and sorting.

A compound index on category then price supports filtering by category and sorting by price efficiently.

🔧 Debug

expert

Why does this MongoDB query run slowly on a read-heavy workload?

Consider this query on a large orders collection:

db.orders.find({ status: "completed" }).sort({ orderDate: -1 }).limit(5)

The query is slow despite indexes on status and orderDate separately. What is the main reason?

A MongoDB cannot use two separate indexes efficiently for filtering and sorting; a compound index is needed.

B The limit(5) clause causes MongoDB to scan the entire collection.

C Indexes on individual fields always slow down queries with sort.

D The query syntax is invalid and causes a full collection scan.

💡 Hint

Think about how MongoDB uses indexes for filtering and sorting together.

MongoDB can only use one index efficiently per query stage. Separate indexes on status and orderDate don't help sorting after filtering. A compound index on {status: 1, orderDate: -1} is needed.