Task 1.2: Transform and Process Data DEA-C01 Practice Questions (Page 3)

Questions 11–15

1. 11foundation · easy

   A data engineer needs to run a complex, long-running ETL job that processes terabytes of data using Spark. Which AWS service is most appropriate?

   • AAWS Glue DataBrew
   • BAmazon EMR
   • CAmazon Redshift
   • DAWS Lambda
   Reveal Solution Discussion Select an answer first
2. 12expert · hard

   A data engineer runs an AWS Glue ETL job that reads JSON files from S3, transforms them, and writes Parquet to another S3 bucket. The job fails with a 'Schema mismatch' error. The engineer inspects the data and finds that some files have an extra field 'customer_rank' that is not present in others. What is the most efficient way to resolve this?

   • AModify the ETL script to use a dynamic frame and apply mapping to handle missing fields.
   • BUse a custom classifier in the Glue crawler to detect the schema.
   • CUse a fixed schema in the ETL script and ignore the extra field.
   • DManually add the missing field to all JSON files before processing.
   Reveal Solution Discussion Select an answer first
4. 13foundation · easy

   Which AWS service can be used to convert data from CSV to Parquet format as part of a serverless ETL pipeline?

   • AAmazon S3
   • BAWS Glue
   • CAmazon Redshift
   • DAWS DataSync
   Reveal Solution Discussion Select an answer first
5. 14foundation · easy

   A data engineer converts a large CSV file to Parquet format for storage in Amazon S3. What is a primary benefit of using Parquet over CSV for analytics?

   • AParquet provides better compression and columnar storage, reducing scan costs
   • BParquet is a schema-less format that requires no schema definition
   • CParquet is a row-based format that is faster for full-row scans
   • DParquet is a plain-text format that is easier to debug
   Reveal Solution Discussion Select an answer first
6. 15expert · hard

   A data engineer is troubleshooting an AWS Glue ETL job that is running slowly. The job reads data from S3, performs a join with a reference table, and writes the results to S3. The job processes 5 GB of data and takes 45 minutes. The engineer notices that the job is using only 2 DPUs. What is the most likely cause of the slow performance?

   • AThe S3 input files are highly compressed, causing CPU overhead.
   • BThe job is not using enough DPUs, so it is under-provisioned.
   • CThe job is using a dynamic frame, which is slower than a data frame.
   • DThe reference table is stored in a format that is not optimized for joins.
   Reveal Solution Discussion Select an answer first