Optimize Performance DP-700 Practice Questions (Page 2)

Questions 6–10

1. 6application · medium

   A data engineer runs a Spark notebook that reads a large Delta table, performs a wide transformation, and writes the result to another Delta table. The job currently takes 45 minutes and uses the default Spark configuration. The cluster is underutilized during the shuffle phase. Which configuration change should the engineer make to reduce job execution time?

   • ADecrease the number of shuffle partitions to reduce overhead.
   • BDisable AQE to avoid plan changes that may cause instability.
   • CIncrease the number of shuffle partitions and enable adaptive query execution (AQE).
   • DUse a single executor with more memory to avoid shuffle overhead.
   Reveal Solution Discussion Select an answer first
2. 7expert · medium

   A data pipeline uses a Copy activity to load data from an on-premises Oracle database to a lakehouse. The pipeline runs every hour and loads the full table. The table has 500 million rows and grows by 5 million rows per day. The pipeline takes 90 minutes, and the source database is under heavy load during the copy. The engineer wants to reduce the load on the source and improve pipeline throughput. Which approach should the engineer take?

   • AImplement incremental loading using a watermark column and enable staging via Azure Blob Storage.
   • BIncrease the degree of copy parallelism to 10 and keep the full load.
   • CSwitch to a Spark notebook activity and use a full load with a large cluster.
   • DUse a Dataflow Gen2 with query folding to push the load to the source.
   Reveal Solution Discussion Select an answer first
4. 8expert · medium

   A Fabric warehouse contains a large fact table with a clustered columnstore index. The table is loaded by a nightly batch job that inserts new rows and updates a small percentage of existing rows. Over time, the columnstore index has become fragmented, and query performance has degraded. The engineer wants to improve query performance without disrupting the nightly load. Which action should the engineer take?

   • ARun ALTER INDEX REORGANIZE on the columnstore index after the nightly load.
   • BDisable the columnstore index and use a rowstore index instead.
   • CDrop the columnstore index and recreate it as a heap.
   • DRun ALTER INDEX REBUILD on the columnstore index every night.
   Reveal Solution Discussion Select an answer first
5. 9application · medium

   A data pipeline uses a Dataflow Gen2 to load data from multiple sources into a lakehouse. The pipeline runs on a schedule and takes longer than expected. The engineer notices that the Dataflow uses a single compute instance and the source queries are not optimized. Which action should the engineer take to improve pipeline throughput?

   • ADisable query folding to force all transformations to run in the Dataflow engine.
   • BEnable staging on the Dataflow and use query folding where possible.
   • CIncrease the number of Dataflow refresh retries to handle failures.
   • DUse a single source query that fetches all data without filtering.
   Reveal Solution Discussion Select an answer first
6. 10foundation · easy

   Which Eventstreams optimization technique involves dividing the event stream into multiple parallel streams to increase processing throughput?

   • AEvent enrichment
   • BEvent deduplication
   • CEvent compression
   • DEvent partitioning
   Reveal Solution Discussion Select an answer first