Unit 5: MongoDB Integration - Practice Quiz

1 What is Mongoose in a Node.js application?

2 Which object is commonly used to define the structure of documents in Mongoose?

3 Which method is commonly used to connect Mongoose to MongoDB?

4 What does a Mongoose model represent?

5 What is PyMongo?

6 Which PyMongo class is commonly used to create a connection to MongoDB?

7 In PyMongo, what does a collection store?

8 Which PyMongo method retrieves a single document?

9 What does the letter C represent in CRUD?

10 Which CRUD operation is used to retrieve documents from MongoDB?

11 Which MongoDB method inserts one document into a collection?

12 Which MongoDB method updates one matching document?

13 Which MongoDB method removes one matching document?

14 What is predictive analytics used for?

15 How can MongoDB support predictive analytics?

16 Which type of data can MongoDB store for an AI prediction system?

17 What is a common purpose of using AI with MongoDB data?

18 What is the main goal of query optimization?

19 What can an index help MongoDB do?

20 What may AI analyze to recommend better database queries?

21 A Mongoose application only needs to read user records and return plain JSON objects. The records will not be modified or saved. Which query is most appropriate for reducing Mongoose document-processing overhead?

22 A developer updates a product with Product.findByIdAndUpdate(id, update). The returned value contains the product data from before the update. Which option should be added to return the updated document?

23 A Mongoose schema requires price to be at least 1. An update using findOneAndUpdate() incorrectly accepts price: 0. Which option should be included in the update operation?

24 An order-creation endpoint must insert an order and decrease product stock as one atomic operation. What is the most appropriate Mongoose approach?

25 A Flask route receives a MongoDB document ID as a URL string. Which conversion is normally required before using it in a PyMongo _id query?

26 A Python service must retrieve only the name and score fields while excluding _id. Which PyMongo query uses the correct projection?

27 A PyMongo application frequently creates a new MongoClient for every request and closes it immediately afterward. What is the preferred design for a typical web service?

28 A data-loading job has 5,000 independent update operations to send through PyMongo. Which approach generally reduces network round trips while allowing the operations to be grouped?

29 An inventory API must reduce stock by 1 only when the current stock is greater than 0. Which update best prevents two concurrent requests from overselling the final item?

30 A REST API calls delete_one({"_id": id}) and receives deleted_count == 0. Which HTTP response is most appropriate when the identifier was valid but no matching resource existed?

31 A profile update endpoint receives only { "city": "Pune" }. The existing name and email fields must remain unchanged. Which update document should the application construct?

32 Two users can edit the same document. The document contains a numeric version field. How should the application implement optimistic concurrency control?

33 A predictive-maintenance model needs each machine's average temperature during the previous hour. Where can this feature be efficiently computed before model inference?

34 A model predicts whether a customer will cancel a subscription tomorrow. Which feature would cause target leakage during training?

35 An application stores model predictions with modelVersion, predictedAt, and featuresVersion. What is the main analytical benefit of storing this metadata?

36 A deployed fraud model should score new transactions shortly after they are inserted into MongoDB. Which MongoDB capability is most suitable for triggering this near-real-time workflow?

37 An AI optimizer observes the frequent query find({ tenantId: T, status: S }).sort({ createdAt: -1 }). Which compound index is the strongest candidate?

38 After an AI system recommends an index, which explain() evidence most strongly indicates that the query became more efficient?

39 An optimizer proposes indexes for every observed query pattern, but the collection has heavy write traffic. Which factor should prevent automatic acceptance of all recommendations?

40 Query traffic changes substantially between business hours and overnight batch processing. What input would best help an AI optimizer adapt its recommendations?

41 A Mongoose schema defines email with unique: true. Two concurrent requests create documents with the same email, and both validations pass before either write completes. What is the reliable way to enforce uniqueness?

42 A Mongoose update uses Model.updateOne({ _id: id }, { $set: payload }), where payload is received from an API client. Which design best prevents unauthorized field changes and operator injection?

43 A service uses findOneAndUpdate() to increment an account balance and expects the returned document to contain the new balance. Which configuration is required for that expectation?

44 A Mongoose query uses populate('customer') and returns thousands of orders. The customer documents contain large profile fields that the API never exposes. Which change most directly reduces response size and hydration cost?

45 A PyMongo application must transfer a newly inserted document's identifier to another collection, and both writes must commit atomically. Which approach is appropriate when the deployment supports transactions?

46 A PyMongo worker retries a transaction after a transient network error. The transaction inserts a document with a client-generated UUID and a unique index on that UUID. Why is this design useful?

47 A PyMongo query sorts by event_time descending and limits results, but performance degrades as the collection grows. The filter is { "tenant_id": t, "status": "open" }. Which index is generally the strongest candidate?

48 A Python API reads documents using find() and converts them to JSON. Some documents contain ObjectId and timezone-aware datetime values. What is the most robust boundary design?

49 An order update must change status from pending to paid only if it is still pending, and the application must know whether the transition occurred. Which operation is most suitable?

50 An application performs pagination with skip(page * 100).limit(100) on a collection receiving continuous inserts. Which issue is most likely, and what is the better strategy for a large collection?

51 A delete endpoint receives a user-supplied filter and passes it directly to deleteMany(). Which control is most important for preventing accidental broad deletion?

52 A bulk import contains 10,000 independent records. Some records violate a unique index, but valid records should still be inserted. Which behavior best matches this requirement?

53 A read-modify-write workflow increments a counter, but concurrent requests occasionally overwrite each other's increments. Which database-side change is the best fix?

54 A predictive model estimates customer churn from historical customer snapshots. The feature pipeline includes each customer's last_support_ticket, but the ticket may have been created after the prediction timestamp. What problem does this create?

55 A feature-generation aggregation joins transactions to accounts and then groups by account. The transaction collection is much larger, and only transactions from the previous 30 days are relevant. Where should the time filter normally be applied?

56 A production model uses a feature that is computed from documents arriving through a change stream. Events can arrive out of order. Which design best preserves correct time-window features?

57 A team evaluates a fraud model by randomly splitting transactions into training and test sets. Fraud patterns evolve over time, and the model will score future transactions. Which evaluation design is more defensible?

58 An optimizer recommends an index { status: 1, region: 1 } for a workload, but the query filters on region and sorts by created_at while omitting status. Why may the recommendation be ineffective?

59 An AI query tuner reports that a query's average latency improved after adding an index, but p99 latency and write throughput both worsened. What is the best conclusion?

60 A learned optimizer detects that a query is slow because a predicate has highly skewed values. It proposes different plans for frequent and rare values. What capability is this exploiting?