IBM Cloud Pak for Data V3.x Data Engineer Intermediate Practice Exam: Medium Difficulty 2025

A data engineer is designing a Cloud Pak for Data deployment where multiple teams need to work with sensitive customer data. The security team requires audit trails of all data access, while different teams need isolated environments. Which architectural approach best meets these requirements?

During a DataStage job execution in Cloud Pak for Data, you notice that a parallel job processing 50 million records is running significantly slower than expected. The job reads from a relational database, performs transformations, and writes to cloud object storage. What is the MOST likely bottleneck to investigate first?

A data engineer needs to create a DataStage flow that consolidates customer data from three different source systems (Oracle, DB2, and REST API) with varying data quality. The business requires that records with missing email addresses be flagged but not rejected, while records with invalid customer IDs must be rejected. What combination of stages should be used?

An organization has implemented IBM Watson Knowledge Catalog and needs to ensure that all data assets containing personally identifiable information (PII) are automatically classified and have appropriate data protection rules applied. What sequence of configuration steps accomplishes this?

A data engineer is building a data pipeline where Data Refinery is used to cleanse and prepare data before it's consumed by a machine learning model. The preparation steps include removing duplicates, standardizing date formats, and creating derived columns. How should this process be productionized for scheduled execution?

When configuring Data Virtualization in Cloud Pak for Data to access multiple heterogeneous data sources including on-premises databases and cloud data lakes, what is the primary consideration for optimizing query performance?

A Cloud Pak for Data environment is experiencing intermittent connection failures to a remote DB2 database used by multiple DataStage jobs. The network team confirms connectivity is stable. What troubleshooting approach should a data engineer take?

An organization needs to establish data quality rules that validate customer addresses across multiple data sources before loading into a master customer database. The rules should check for completeness, format consistency, and cross-reference with a postal code database. Which Cloud Pak for Data capability best addresses this requirement?

A data engineer needs to create a virtualized view that joins customer data from an on-premises Oracle database with transaction data from a cloud-based PostgreSQL database. The business requires near real-time access but reports that queries are timing out. What optimization strategy should be implemented?

A Cloud Pak for Data project team notices that their DataStage jobs that previously ran successfully are now failing with 'insufficient memory' errors. The jobs process the same data volumes as before, and no code changes were made. What is the MOST systematic approach to diagnose and resolve this issue?