Unit 6: Advance Databases - Practice Quiz

CAP570 — Advanced Database Techniques 60 Questions
0 Correct 0 Wrong 60 Left
0/60

1 Which database trend uses artificial intelligence to automate tasks such as tuning and optimization?

Current Trends and Future Directions in Database Technology Easy
A. Flat-file databases
B. Autonomous databases
C. Manual databases
D. Desktop databases

2 What is a major goal of modern database technology?

Current Trends and Future Directions in Database Technology Easy
A. Avoiding data analysis
B. Improving scalability
C. Removing all networks
D. Reducing all data

3 Which technology allows database processing to occur closer to where data is generated?

Current Trends and Future Directions in Database Technology Easy
A. Edge computing
B. Batch printing
C. Manual indexing
D. Tape archiving

4 Which characteristic of Big Data refers to the enormous amount of data being generated?

Big Data and Databases Easy
A. Volume
B. Veracity
C. Variety
D. Velocity

5 Which Big Data characteristic describes the speed at which data is generated and processed?

Big Data and Databases Easy
A. Veracity
B. Variety
C. Velocity
D. Volume

6 Which Big Data characteristic refers to structured, semi-structured, and unstructured data formats?

Big Data and Databases Easy
A. Volume
B. Variety
C. Velocity
D. Veracity

7 Which framework is commonly associated with distributed storage and processing of Big Data?

Big Data and Databases Easy
A. HTML
B. SQLite
C. FTP
D. Hadoop

8 What does OLAP stand for?

OLAP Easy
A. Online Analytical Processing
B. Operational Application Processing
C. Offline Analytical Protocol
D. Online Application Programming

9 What is the primary purpose of an OLAP system?

OLAP Easy
A. Analyzing business data
B. Editing source code
C. Designing web pages
D. Managing network cables

10 Which OLAP operation selects a single value from one dimension of a data cube?

OLAP Easy
A. Drill-down
B. Pivot
C. Roll-up
D. Slice

11 Which OLAP operation moves from summarized data to more detailed data?

OLAP Easy
A. Roll-up
B. Slice
C. Rotate
D. Drill-down

12 Where is a cloud database typically hosted?

Cloud Databases Easy
A. On paper records
B. Inside source code
C. Inside a spreadsheet
D. On cloud infrastructure

13 Which is a common advantage of cloud databases?

Cloud Databases Easy
A. Manual availability
B. Fixed capacity
C. Offline-only access
D. Elastic scalability

14 What does DBaaS stand for in cloud computing?

Cloud Databases Easy
A. Database as a System
B. Database as a Service
C. Data Browser and Security
D. Data Backup as Storage

15 What is a main goal of NewSQL databases?

NewSQL Easy
A. Removing support for transactions
B. Replacing tables with paper files
C. Combining SQL with horizontal scalability
D. Restricting databases to one user

16 Which property is commonly preserved by NewSQL databases?

NewSQL Easy
A. ACID transactions
B. Schema-free storage
C. Unverified records
D. Manual replication

17 NewSQL databases are most closely related to which traditional database model?

NewSQL Easy
A. Relational model
B. Graph model
C. Document model
D. Key-value model

18 What does NoSQL commonly mean?

NoSQL Easy
A. New Online SQL
B. No Structured Queries
C. Not Only SQL
D. Network-Oriented SQL

19 Which NoSQL database type stores information as pairs containing a key and its associated value?

NoSQL Easy
A. Key-value database
B. Graph database
C. Document database
D. Column-family database

20 Which NoSQL database type is especially suitable for representing relationships between entities?

NoSQL Easy
A. Column-family database
B. Key-value database
C. Document database
D. Graph database

21 A retailer wants its database to automatically identify unusual purchasing patterns and recommend indexes as workloads change. Which emerging trend best supports this requirement?

Current Trends and Future Directions in Database Technology Medium
A. Static query compilation
B. Manual database normalization
C. AI-assisted database management
D. File-based data processing

22 An application must process data near thousands of industrial sensors because sending every event to a central server causes excessive delay. Which database trend is most appropriate?

Current Trends and Future Directions in Database Technology Medium
A. Single-server warehousing
B. Centralized batch processing
C. Manual data archiving
D. Edge database processing

23 A company needs to keep an auditable record that multiple organizations can verify without relying on one central database owner. Which technology is the best fit?

Current Trends and Future Directions in Database Technology Medium
A. Distributed ledger database
B. Centralized document database
C. In-memory relational database
D. Local embedded database

24 A streaming platform stores petabytes of click events across hundreds of machines. Adding more machines should increase storage and processing capacity. Which design principle is being applied?

Big Data and Databases Medium
A. Horizontal scalability
B. Transaction serialization
C. Schema normalization
D. Vertical scalability

25 A sensor platform receives millions of readings per second and must detect faults within seconds. Which big-data characteristic creates the main challenge?

Big Data and Databases Medium
A. Velocity
B. Volume
C. Veracity
D. Variety

26 A data pipeline must combine relational sales tables, JSON web logs, images, and free-text reviews. Which big-data characteristic is most directly represented?

Big Data and Databases Medium
A. Volatility
B. Variety
C. Velocity
D. Validity

27 A distributed analytics job performs the same calculation independently on many data partitions and then combines the partial results. Which processing pattern does this describe?

Big Data and Databases Medium
A. Two-phase locking
B. MapReduce
C. Referential integrity
D. Write-ahead logging

28 An analyst is viewing annual revenue by country and wants to display monthly revenue for individual cities. Which OLAP operation is required?

OLAP Medium
A. Roll-up
B. Slice
C. Drill-down
D. Pivot

29 A sales cube contains the dimensions Product, Region, and Time. An analyst fixes Time to 2025 while examining all products and regions. Which OLAP operation is this?

OLAP Medium
A. Drill-through
B. Dice
C. Slice
D. Roll-up

30 A company frequently calculates total sales by month, product category, and region. Which schema is typically simplest for these OLAP queries?

OLAP Medium
A. Key-value schema
B. Document schema
C. Graph schema
D. Star schema

31 An analyst changes a report from displaying regions as rows and years as columns to displaying years as rows and regions as columns. Which operation was performed?

OLAP Medium
A. Drill-down
B. Pivot
C. Slice
D. Roll-up

32 An online store experiences unpredictable traffic spikes and wants database capacity to increase and decrease automatically. Which cloud capability is most relevant?

Cloud Databases Medium
A. Schema denormalization
B. Manual replication
C. Elastic scaling
D. Static partitioning

33 A development team wants the cloud provider to handle database patching, backups, and routine availability management. Which service model should it choose?

Cloud Databases Medium
A. Managed database service
B. Local database server
C. Self-hosted virtual machine
D. Shared network storage

34 A company deploys database replicas in three cloud availability zones. What is the primary purpose of this design?

Cloud Databases Medium
A. Simplified query syntax
B. Reduced schema complexity
C. Stronger data compression
D. Improved fault tolerance

35 A payment platform requires SQL, ACID transactions, and horizontal scaling across many nodes. Which database category best matches these requirements?

NewSQL Medium
A. NewSQL
B. Key-value NoSQL
C. Single-node OLAP
D. Traditional file system

36 A distributed SQL database partitions customer rows across nodes while preserving transactional consistency. Which NewSQL technique is primarily responsible for distributing the workload?

NewSQL Medium
A. Automatic sharding
B. Log compression
C. Local caching
D. Column normalization

37 Why might a bank select NewSQL instead of a typical eventually consistent NoSQL database for account transfers?

NewSQL Medium
A. It stores only unstructured documents
B. It supports distributed ACID transactions
C. It avoids transaction coordination
D. It eliminates all database schemas

38 A shopping application stores product records whose attributes differ widely by category and change frequently. Which NoSQL model is most suitable?

NoSQL Medium
A. Graph database
B. Key-value cache
C. Document database
D. Wide-column database

39 A social network must efficiently find friends of friends and the shortest connection path between users. Which NoSQL model is the best fit?

NoSQL Medium
A. Key-value database
B. Wide-column database
C. Graph database
D. Document database

40 A distributed NoSQL system continues accepting updates during a network partition, but replicas may temporarily return different values. Which CAP trade-off does this illustrate?

NoSQL Medium
A. Durability over isolation
B. Availability over consistency
C. Atomicity over partition tolerance
D. Consistency over availability

41 An HTAP system must run long analytical scans over rapidly changing operational data without blocking transactions, while keeping analytical results nearly current. Which design most directly satisfies both requirements?

Current Trends and Future Directions in Database Technology Hard
A. Copy the operational database nightly into a separate analytical database
B. Use MVCC snapshots with incremental propagation to an analytical execution engine
C. Replace transactional indexes with compressed column-oriented storage on the primary
D. Run all analytical queries under serializable locking on the primary engine

42 A vector database searches 500 million embeddings, but every query also includes a highly selective tenant predicate matching only of rows. A global ANN search followed by predicate filtering often returns fewer than results. Which strategy best addresses the problem?

Current Trends and Future Directions in Database Technology Hard
A. Increase embedding dimensionality before performing the global ANN search
B. Use filter-aware ANN or prefilter candidates through tenant-partitioned indexes
C. Replace cosine similarity with Euclidean distance for all tenant searches
D. Apply the tenant predicate after ANN and reduce the requested value of

43 A learned index models the cumulative distribution of sorted keys to predict record positions. Which mechanism is still necessary for exact lookup when the model has prediction error or encounters distribution drift?

Current Trends and Future Directions in Database Technology Hard
A. A Bloom filter that directly returns the exact position of every key
B. A larger neural model that eliminates the need to compare stored keys
C. A bounded local search plus a fallback structure with correctness guarantees
D. A periodic cache flush that restores the original key distribution

44 A distributed aggregation groups ten billion records by customer ID, but one customer owns of all records and causes a single reducer to dominate runtime. Which exact transformation best mitigates this skew?

Big Data and Databases Hard
A. Replicate every input partition to several reducers and retain the fastest result
B. Range-partition customer IDs and assign the hot customer to the first reducer
C. Salt the hot key, partially aggregate each salted group, then merge its partial results
D. Sort all records globally by customer ID before sending them to one reducer

45 A stateful stream processor may fail after updating its state but before acknowledging source records. Which design provides end-to-end exactly-once effects when the sink supports transactions?

Big Data and Databases Hard
A. Acknowledge source records before processing and asynchronously update the sink
B. Replay unacknowledged records and assume that duplicate sink writes are harmless
C. Checkpoint operator state independently while committing every sink write immediately
D. Atomically checkpoint source offsets and state, then transactionally commit sink output

46 A data lake stores clickstream events. Most queries select three of fifty columns and restrict event_date to a narrow range. Which physical design most effectively reduces both file scanning and column I/O?

Big Data and Databases Hard
A. Partition by every projected column and compress complete rows together
B. Partition by random event UUID and store each partition in row format
C. Partition by event_date and store each partition in a columnar format
D. Store one globally sorted text file and build only an in-memory hash table

47 In a sales cube, a product may belong to multiple promotional categories simultaneously. Summing category totals therefore exceeds the product-level total. Which dimensional design preserves summarizability?

OLAP Hard
A. Choose an arbitrary category during cube refresh and discard other memberships
B. Convert category membership into a degenerate dimension inside the fact table
C. Store each product under every category and apply ordinary additive aggregation
D. Use a bridge table with allocation weights whose per-product sum equals

48 A cube has four dimensions containing respectively 3, 2, 4, and 1 non-ALL hierarchy levels. If a cuboid chooses exactly one level or ALL from each dimension, how many distinct cuboids are possible?

OLAP Hard
A.
B.
C.
D.

49 A monthly snapshot fact table stores each account's end-of-day balance. Which aggregation is semantically valid for computing a month-end balance by branch?

OLAP Hard
A. Select each branch's largest daily balance, then sum those branch maxima
B. Select each account's final daily balance, then sum those balances by branch
C. Sum every daily balance for each account, then sum the results by branch
D. Average all account-day balances, then multiply by the number of accounts

50 A warehouse uses a Type 2 slowly changing customer dimension. A late-arriving sale occurred before a customer's region changed but is loaded afterward. How should the fact be linked?

OLAP Hard
A. Match the sale timestamp to the dimension version's effective interval
B. Link the sale to the customer's currently active dimension version
C. Create a new customer version whose effective date equals the load timestamp
D. Link the sale to every historical customer version and divide its measure

51 A cloud analytical platform separates compute from shared object storage. Two departments require workload isolation but must query the same current tables without creating full data copies. Which approach best exploits this architecture?

Cloud Databases Hard
A. Assign separate object stores and synchronize them through nightly exports
B. Run independent compute warehouses over shared, versioned storage metadata
C. Place both departments in one warehouse and prioritize queries by arrival time
D. Replicate every table into department-specific local disks before each query

52 A multi-region database acknowledges a write after storing it only in Region A. The requirement is to lose no acknowledged writes even if Region A is permanently destroyed immediately afterward. What change is necessary?

Cloud Databases Hard
A. Asynchronously copy the write to another region after returning acknowledgment
B. Increase the number of replicas within Region A before returning acknowledgment
C. Synchronously replicate the write to another failure-independent region before acknowledgment
D. Cache the acknowledged write at the client until the next regional backup

53 A managed database takes a full backup every 24 hours and durably archives transaction-log segments every 5 minutes. Assuming restoration can replay all archived logs, what is the worst-case recovery point objective after a database failure?

Cloud Databases Hard
A. Approximately zero minutes
B. Approximately 24 hours
C. Approximately 5 minutes
D. Approximately 24 hours and 5 minutes

54 A deterministic NewSQL engine globally orders transactions before execution to reduce runtime coordination. Which workload characteristic most directly weakens the benefit of this approach?

NewSQL Hard
A. Transactions that execute stored procedures over known partitioning keys
B. Transactions whose keys and operations are declared before global sequencing
C. Transactions that perform conflict-free writes to disjoint record ranges
D. Transactions whose read and write sets cannot be determined before execution

55 A globally distributed NewSQL system assigns a commit timestamp using a bounded clock-uncertainty interval. Why might it delay acknowledging the commit until the uncertainty interval has passed?

NewSQL Hard
A. To guarantee that every replica has compacted its complete storage history
B. To eliminate the need for consensus when replicating the transaction log
C. To ensure later transactions cannot appear to commit before this transaction
D. To ensure that the transaction accessed records from only one physical shard

56 A distributed SQL table is range-sharded by an ever-increasing timestamp. Nearly all inserts target the newest shard, producing a write hotspot. Which redesign best distributes writes, and what trade-off does it introduce?

NewSQL Hard
A. Create smaller newest ranges; inserts remain localized but use fewer replicas
B. Use a larger timestamp type; time-range queries may require additional casting
C. Add a hash-derived prefix; time-range queries may need to scan multiple shards
D. Replicate the newest shard; writes become local to every replica simultaneously

57 A Dynamo-style store uses replication factor , read quorum , and write quorum . Which statement is most accurate?

NoSQL Hard
A. Every read is linearizable because both and exceed half of
B. Writes remain available under any three replica failures because
C. Read and write quorums cannot intersect unless is at least
D. Ordinary quorums intersect because , but sloppy quorums can weaken that guarantee

58 An order and its bounded set of order lines are almost always read and updated together. Cross-order queries are rare. Which document-model design best matches this access pattern?

NoSQL Hard
A. Embed order lines inside the order document and update them atomically
B. Store every line in a separate collection and join them during each order read
C. Embed all customer orders inside one customer document without size limits
D. Store order lines as unrelated key-value entries with randomly generated keys

59 A fraud-detection workload repeatedly evaluates variable-length paths such as account → device → account → merchant, with predicates on intermediate relationships. Which database model most naturally supports this workload?

NoSQL Hard
A. A key-value database that scans every key for each traversal step
B. A wide-column database with one partition containing all global relationships
C. A property graph database with indexed starts and adjacency-based traversal
D. A document database that duplicates the complete graph into every account

60 In an eventually consistent LSM-tree store, a deleted value unexpectedly reappears after an offline replica rejoins. Which configuration error most plausibly caused this resurrection?

NoSQL Hard
A. Compaction merged overlapping sorted tables while retaining the newest version
B. Memtables were flushed before reaching their configured maximum capacity
C. Tombstones were purged before all replicas received or repaired the deletion
D. Bloom filters produced a false positive while checking an immutable table