A Spark job aggregates data by a key that is heavily skewed. The engineer tried increasing spark.sql.shuffle.partitions, but the job is still slow because a few tasks process most of the data. The engineer wants to balance the load without changing the final result. Which approach should the engineer take?

Use a broadcast join to avoid the shuffle.

Use repartitionByRange on the key.

Use repartition with a custom partitioner that distributes keys based on their frequency.

Use coalesce to reduce the number of partitions.

A data engineer loads a large DataFrame from Parquet into memory and notices that one executor processes significantly more data than the others, causing the job to run slower than expected. The DataFrame has 200 partitions, but the data is unevenly distributed across keys. The engineer wants to rebalance the data across partitions without increasing the total number of partitions. Which action should the engineer take?

Call repartition(100) on the DataFrame.

Call coalesce(200) on the DataFrame.

Set spark.sql.shuffle.partitions to 200.

Call repartition(200) on the DataFrame.

DatabricksCertified Associate Developer for Apache Spark

Domain 1Objective 5

Configure Spark Partitioning in Distributed Data Processing, Including Shuffles and Partitions ASSOCIATE-DEVELOPER-APACHE-SPARK Practice Questions (Page 6)

Part of the Apache Spark Architecture and Components domain, which accounts for 20% of the ASSOCIATE-DEVELOPER-APACHE-SPARK exam. Databricks does not publish an official question count, but from its 90-minute exam (~35–60 total, ~7–12 in this domain), expect 1–2 from this objective — we provide 29 practice questions to prepare you well beyond it. (estimate)

29questions here

6free pages

8concepts

20%of the exam

Questions 26–29

26foundation · easy
Which of the following operations typically triggers a shuffle?
Select an answer first
27expert · hard
A Spark job aggregates data by a key that is heavily skewed. The engineer tried increasing spark.sql.shuffle.partitions, but the job is still slow because a few tasks process most of the data. The engineer wants to balance the load without changing the final result. Which approach should the engineer take?
Select an answer first
28application · medium
A data engineer loads a large DataFrame from Parquet into memory and notices that one executor processes significantly more data than the others, causing the job to run slower than expected. The DataFrame has 200 partitions, but the data is unevenly distributed across keys. The engineer wants to rebalance the data across partitions without increasing the total number of partitions. Which action should the engineer take?
Select an answer first
29application · medium
A Spark SQL query joins two large tables and then aggregates the result. The cluster has 100 cores. The engineer wants to minimize the number of tasks while keeping all cores busy. What is the most appropriate value for spark.sql.shuffle.partitions?
Select an answer first

Finished these 4 questions?

Review the revealed explanations, or continue through the curriculum.

No more pagesBack to ASSOCIATE-DEVELOPER-APACHE-SPARK

Free Basic Practice is a study aid with revealable answers — not a scored exam. Examers.io is independent and not affiliated with or endorsed by Databricks. “ASSOCIATE-DEVELOPER-APACHE-SPARK” is a trademark of its owner, used for identification only.

Configure Spark Partitioning in Distributed Data Processing, Including Shuffles and Partitions ASSOCIATE-DEVELOPER-APACHE-SPARK Practice Questions (Page 6)

Questions 26–29

Which of the following operations typically triggers a shuffle?

A Spark SQL query joins two large tables and then aggregates the result. The cluster has 100 cores. The engineer wants to minimize the number of tasks while keeping all cores busy. What is the most appropriate value for spark.sql.shuffle.partitions?

Keep studying — Apache Spark Architecture and Components