IBM Certification Big Data Architect 5 — Questions and Answers
Question 1: Which IBM data architecture pattern uses a 'data vault' modeling approach to improve auditability and adaptability in enterprise data warehouses?
- Star schema with conformed dimensions
- Data Vault 2.0 with hubs, links, and satellites (Correct answer)
- Third normal form relational modeling
- Wide denormalized flat tables
Correct answer: Data Vault 2.0 with hubs, links, and satellites
Data Vault 2.0 organizes data into hubs (business keys), links (relationships), and satellites (context/attributes), enabling auditability and schema flexibility.
Question 2: An architect designs an IBM data pipeline where Apache Spark jobs process HDFS data. What is the performance advantage of using Spark's in-memory processing over MapReduce?
- Spark writes intermediate results to disk after each stage like MapReduce
- Spark keeps intermediate data in memory across stages, reducing disk I/O significantly (Correct answer)
- Spark uses fewer CPU cores by limiting parallelism
- Spark avoids YARN entirely by using its own resource manager
Correct answer: Spark keeps intermediate data in memory across stages, reducing disk I/O significantly
Spark caches intermediate results in memory across transformation stages, dramatically reducing the disk I/O that MapReduce incurs between each map and reduce step.
Question 3: In IBM Cloud Object Storage, what erasure coding configuration provides the best balance between storage efficiency and fault tolerance?
- 3x replication across all buckets
- 10+4 erasure coding (10 data slices, 4 parity slices) (Correct answer)
- Single-region replication without redundancy
- RAID-5 on-premise disk configuration
Correct answer: 10+4 erasure coding (10 data slices, 4 parity slices)
10+4 erasure coding stores data across 14 slices with 4 parity slices, tolerating 4 simultaneous failures while using less storage than 3x replication.
Question 4: When integrating IBM MQ with a big data pipeline, what problem does persistent message delivery primarily solve?
- Compressing large binary payloads before HDFS ingestion
- Guaranteeing message delivery even if downstream consumers are temporarily unavailable (Correct answer)
- Encrypting messages in transit between Kafka and Spark
- Deduplicating records before landing in the data lake
Correct answer: Guaranteeing message delivery even if downstream consumers are temporarily unavailable
IBM MQ's persistent delivery ensures messages survive broker restarts and consumer outages, providing at-least-once delivery guarantees for critical data pipelines.
Question 5: A Big Data Architect must enforce column-level security on an IBM Db2 Warehouse table containing PII. Which mechanism is most appropriate?
- Row permissions using IBM row and column access control (RCAC) (Correct answer)
- Table-level GRANT statements to individual users
- Encrypting the entire tablespace with AES-256
- Partitioning PII columns into a separate table
Correct answer: Row permissions using IBM row and column access control (RCAC)
IBM Db2 Row and Column Access Control (RCAC) allows fine-grained policies that restrict which rows and columns specific users or roles can access.
Question 6: Which IBM tool provides automated discovery of sensitive data (PII, PHI, PCI) across heterogeneous big data sources to support compliance requirements?
- IBM Watson Knowledge Catalog with data discovery rules (Correct answer)
- IBM Cognos Controller
- IBM Planning Analytics
- IBM Rational Rhapsody
Correct answer: IBM Watson Knowledge Catalog with data discovery rules
IBM Watson Knowledge Catalog includes automated data discovery with configurable rules that identify and classify sensitive data across connected sources.
Question 7: In designing an IBM big data solution for a financial institution, why is data lineage tracking considered a critical architectural requirement?
- It reduces the storage footprint of raw data in the data lake
- It enables auditors to trace how any data point was derived and transformed end-to-end (Correct answer)
- It improves query performance by caching frequently accessed lineage paths
- It eliminates the need for data quality checks at ingestion
Correct answer: It enables auditors to trace how any data point was derived and transformed end-to-end
Data lineage provides a complete audit trail of data origin, transformations, and movement, which is essential for regulatory compliance and impact analysis.
IBM Certification Big Data Architect 5
Which IBM data architecture pattern uses a 'data vault' modeling approach to improve auditability and adaptability in enterprise data warehouses?