Unit 6: Project - Practice Quiz

1 What is the first step when building a NoSQL mini project?

2 Why is sample data used during mini-project development?

3 Which operation adds a new document to a NoSQL database?

4 What is the main purpose of a database schema or data model in a mini project?

5 Which practice helps prevent unauthorized access to a production database?

6 Why are regular database backups important in production?

7 What does database monitoring help a production team identify?

8 Which principle gives users only the permissions required for their tasks?

9 What is a viva in an academic course?

10 What does a final assessment mainly evaluate?

11 Which preparation method is most useful before a project viva?

12 What is a quiz commonly used to check?

13 What is the main purpose of a course recap?

14 Which feature is commonly associated with NoSQL databases?

15 Why is class discussion useful during a course recap?

16 What does CRUD stand for?

17 What is an AI-augmented database?

18 What is the main goal of AutoML?

19 Which database task can AI help automate?

20 How can anomaly detection assist database administrators?

21 A mini project stores customer profiles in a document database. Which design best supports frequent updates to a customer's contact information?

22 A project dashboard frequently retrieves orders by customer_id and sorts them by created_at. Which index is most appropriate?

23 A mini project must allow users to add new product attributes without changing a fixed table schema. Which NoSQL model is most suitable?

24 A team is choosing between embedding and referencing order items. Order items are usually retrieved only with their order and are rarely shared. Which choice is generally better?

25 An application must remain available when one database node fails. Which production practice most directly addresses this requirement?

26 A production NoSQL query becomes slow after the dataset grows substantially. What should be done first?

27 Which backup strategy provides the strongest protection against accidental deletion and regional infrastructure failure?

28 A service uses an eventually consistent read immediately after writing a user's email address. What outcome should the application be prepared to handle?

29 Which metric is most useful for detecting that a production NoSQL cluster is approaching capacity?

30 During a viva, a student is asked why denormalization is common in NoSQL databases. Which answer is most accurate?

31 A quiz asks which situation is most appropriate for a graph database. Which scenario should be selected?

32 A project assessment requires evidence that an API handles duplicate requests safely. Which test best evaluates this property?

33 Which explanation would best justify selecting a key-value database for a shopping-cart service?

34 A team chooses a NoSQL database mainly because its workload requires horizontal scaling and flexible records. Which concept supports this decision?

35 During a course discussion, a team reports that a partition key creates an uneven distribution of records. What problem does this indicate?

36 Which statement best summarizes the trade-off between strong and eventual consistency?

37 A team reviews a NoSQL design and discovers that its data model follows table structure instead of application queries. What improvement should be discussed?

38 An AI-augmented database detects that a query is repeatedly slow and recommends a new index. What should an administrator do before applying it automatically?

39 A company uses AutoML to predict customer churn from NoSQL records. Which step is essential before training the model?

40 An AI system generates vector embeddings for product descriptions. What database capability is needed to support semantic product search?

41 A mini-project must ingest millions of immutable IoT readings while serving the latest state of each device with single-digit-millisecond latency. Which data model best separates these conflicting access patterns?

42 A recommendation project trains a model from NoSQL interaction data and computes features again during online inference. Offline accuracy is high, but production accuracy is poor because feature values differ between training and serving. Which design most directly prevents this failure?

43 A fraud-detection mini-project stores customer profiles, transactions, and relationships. It requires frequent variable-depth traversals to identify circular money flows, while profile retrieval remains document-oriented. Which architecture is most appropriate?

44 A team claims its mini-project will sustain 20,000 writes per second after observing that rate from uniformly distributed synthetic keys on a single-node test environment. Which evaluation provides the strongest evidence for the claim?

45 A production document database must rename field phone to phoneNumber without downtime while old and new application versions overlap. Which migration sequence is safest?

46 An orders collection is partitioned by tenantId. One large tenant generates 70% of all writes, creating a hot partition, but queries usually filter by tenant and recent time range. Which redesign best addresses the hotspot?

47 A message consumer writes payment results to a NoSQL database. It may crash after the database commit but before acknowledging the message, causing redelivery. Which mechanism most reliably prevents duplicate payment effects?

48 A replicated NoSQL cluster has hourly snapshots and continuous write-ahead-log archiving. The stated recovery point objective is five minutes. Which practice most convincingly verifies that objective?

49 During a network partition, a NoSQL service must accept writes in both regions and cannot coordinate synchronously. What consequence should a strong viva answer identify?

50 A query filters on country, sorts by createdAt descending, and returns the newest 20 records. country has low cardinality, while createdAt is highly selective by range. Which compound index most directly supports filtering and ordered retrieval?

51 In a viva, a student says, "Eventual consistency means every replica becomes identical after a fixed number of seconds." Which correction is most accurate?

52 A quiz presents two replicated counters updated concurrently from initial value 10. Region A increments by 3, while Region B increments by 5. A last-write-wins register stores only the resulting absolute values. What failure can occur after reconciliation?

53 A globally distributed database remains fully operational without network partitions. It can either synchronously coordinate writes for lower staleness or avoid coordination for lower latency. Which principle best captures this trade-off?

54 An e-commerce document model embeds all product details in every order. Product names must reflect their value at purchase time, but a product's current safety-recall status must update immediately across all orders. Which redesign best preserves both semantics?

55 Multiple offline replicas must increment and decrement an inventory adjustment value, then merge without losing concurrent operations. Assuming every operation is eventually delivered, which data type is most suitable?

56 A fraud model consumes a change-data-capture stream from a NoSQL database. After a schema migration, the field amount becomes transaction.amount, and predictions silently degrade. Which course-level lesson best addresses this failure?

57 A learned index predicts record positions accurately under the training distribution, but incoming keys later become highly skewed. Which production design best limits worst-case lookup degradation?

58 An AutoML pipeline predicts customer churn using historical NoSQL documents. Each document is later updated with retentionOfferAccepted, which is known only after the prediction date. Cross-validation reports unrealistically high accuracy. What is the best correction?

59 An AI database assistant generates executable NoSQL queries from natural language. It occasionally proposes unrestricted updates when users ask ambiguous questions. Which control offers the strongest production safeguard?

60 A vector search system retrieves semantically similar documents but often returns records from unauthorized tenants. The team also needs high relevance within the permitted tenant. Which pipeline is most appropriate?