Google Professional Data Engineer
Get started today
Ultimate access to all questions.
Ultimate access to all questions.
As the architect of a seismic data analysis system utilizing an Apache Hadoop cluster for its ETL process, you face a challenge where the ETL process takes several days to complete due to computational complexity. Upon investigation, you find that the sensor calibration step was omitted. To systematically incorporate sensor calibration in the future, what should you do?
Explanation:
To ensure systematic sensor calibration, introducing a new MapReduce job dedicated to applying calibration to raw data as the first step in the ETL process is the most effective approach. This guarantees that all subsequent jobs process accurately calibrated data, enhancing efficiency. Option A's predictive correction may lack accuracy. Option B's manual calibration invites inconsistency. Option D risks processing uncalibrated data in initial steps, leading to inaccuracies.