CDS Data Management & Integration 3 — Questions and Answers
Question 1: In an ETL pipeline, at which stage should data cleansing and standardization typically occur?
- Extract
- Transform (Correct answer)
- Load
- Query
Correct answer: Transform
The Transform stage is where data is cleansed, standardized, enriched, and restructured before being loaded into the target system.
Question 2: A company's data warehouse team receives source data where the 'customer_id' field in the CRM uses integers but the ERP uses alphanumeric codes for the same entity. This mismatch is called a:
- Schema drift
- Key reconciliation problem (Correct answer)
- Data latency issue
- Cardinality mismatch
Correct answer: Key reconciliation problem
Key reconciliation involves resolving differences in how the same real-world entity is identified across different source systems.
Question 3: Which of the following BEST describes an ELT (Extract, Load, Transform) approach versus traditional ETL?
- ELT performs transformations before loading into the target
- ELT loads raw data first and performs transformations within the target system (Correct answer)
- ELT is used exclusively for unstructured data
- ELT eliminates the need for data governance
Correct answer: ELT loads raw data first and performs transformations within the target system
ELT leverages the processing power of modern cloud data warehouses by loading raw data first and transforming it in place using SQL or native tools.
Question 4: What is the role of a 'data steward' versus a 'data owner' in data governance?
- Data stewards own data assets; data owners manage day-to-day quality
- Data owners are accountable for data assets; data stewards manage day-to-day quality and compliance (Correct answer)
- Both roles have identical responsibilities
- Data stewards approve data access; data owners approve data definitions
Correct answer: Data owners are accountable for data assets; data stewards manage day-to-day quality and compliance
Data owners hold business accountability for data assets, while data stewards are responsible for the operational management of data quality, definitions, and usage.
Question 5: Which concept describes the practice of storing a copy of data from operational systems in a staging area before loading it into the data warehouse?
- Data virtualization
- Data federation
- Data staging (Correct answer)
- Data masking
Correct answer: Data staging
A data staging area holds extracted data temporarily, decoupling the extraction from transformation and protecting source systems from heavy query loads.
Question 6: In the context of data integration, what is 'schema mapping'?
- Creating visual diagrams of database server hardware
- Defining the correspondence between fields in different data schemas (Correct answer)
- Encrypting schema metadata for security
- Automatically generating database indexes
Correct answer: Defining the correspondence between fields in different data schemas
Schema mapping identifies how data fields in a source schema correspond to fields in a target schema to enable accurate data transformation.
Question 7: A retailer wants to analyze sales trends without impacting the performance of its transactional POS system. The BEST architectural decision is to:
- Run analytical queries directly against the POS database during off-peak hours
- Replicate data into a separate analytical data store (Correct answer)
- Archive old POS data to CSV files for analysis
- Add more CPU cores to the POS server
Correct answer: Replicate data into a separate analytical data store
Separating analytical workloads into a dedicated data warehouse or data mart prevents OLAP queries from degrading the OLTP system performance.
CDS Data Management & Integration 3
In an ETL pipeline, at which stage should data cleansing and standardization typically occur?