Free Cloudera CDP-3002 Exam Dumps Questions & Answers

You're working with an Airflow DAG that performs data quality checks on sensitive dat a. How can you ensure data security during the checks?

• A.  All of the above
• B.  Implement data masking techniques to obfuscate sensitive information during the checks.
• C.  Utilize environment variables to store sensitive data and access them within the PythonOperator.
• D.  Store sensitive data like thresholds and comparison values directly within the DAG code.

Which of the following is a benefit of using broadcast variables in Spark for caching static lookup tables?

• A.  They are automatically cleaned up after each task.
• B.  They reduce network I/O by making data available locally on each node.
• C.  They reduce the reliability of Spark applications.
• D.  They increase the amount of data shuffle during joins.

You're working with a large dataset that needs to be partitioned and processed in chunks to improve efficiency. How can you achieve this using Airflow operators?

• A.  Leverage the File transform operator to partition the data based on specific criteria within the operator itself.
• B.  Use the Split Operator to divide the data into smaller subsets and chain them with downstream processing tasks.
• C.  Implement a custom Python script to handle partitioning and then use the BashOperator to execute the script within the DAG.
• D.  Configure the source system to provide the data pre-partitioned for efficient processing.

You want to schedule your Airflow DAG to run every hour, starting at midnight (00:00). How can you achieve this scheduling configuration?

• A.  Set the schedule_interval parameter to "0 0 ".
• B.  Use the cron expression "@hourly".
• C.  Define a custom Python function to schedule the DAG at the desired time.
• D.  Utilize Airflow triggers to initiate the DAG execution at midnight.

What technique can be employed to optimize join performance by reducing data shuffle across the network?

• A.  Decreasing the memory allocated to executors
• B.  Using broadcast joins for smaller datasets
• C.  Partitioning both datasets on a different key
• D.  Increasing the number of partitions