Free Databricks-Certified-Data-Engineer-Associate: Databricks Certified Data Engineer Associate Exam Questions and Answers

Another team's pipeline updates the Unity Catalog table main.sales.orders_silver at irregular times during the day. A data engineer needs a Lakeflow Job that refreshes a downstream aggregate each time that table receives new data, without polling on a fixed interval. Which trigger should the data engineer configure on the job?

• AA. Continuous trigger on the job
• BB. Scheduled trigger with a cron expression that runs every 15 minutes
• CC. File arrival trigger on the storage location of main.sales.orders_silver
• DD. Table update trigger on main.sales.orders_silver

A data engineer opens the Spark UI for a slow join. Stage 7 has 200 tasks. The median task duration is 8 seconds and the median shuffle read is 120 MB. One task has run for 16 minutes and shows a shuffle read of 38 GB, while the remaining tasks have finished. What is the most likely cause?

• AA. Data skew, where one join key value produces a disproportionately large partition
• BB. Driver memory exhaustion while results are collected from the executors
• CC. Slow storage reads caused by a large number of small input files in the source table
• DD. Too many shuffle partitions, which creates scheduling overhead for very small tasks

A Lakeflow Job has five tasks that run in sequence. The first three tasks succeeded and the fourth task failed because of a missing source file. After the file is restored, the data engineer wants to rerun the failed task and the tasks that depend on it without rerunning the successful tasks. What should the data engineer use?

• AA. A repair run that selects the failed task and its dependent tasks
• BB. A retry policy on the failed task set to three retries
• CC. Run now with different job parameters
• DD. A clone of the job, followed by a manual run of the clone

A data engineer joins a 3 TB table named orders with a dimension table named regions whose estimated size is 40 MB. The Spark UI shows a sort-merge join with a large shuffle of the orders table. The spark.sql.autoBroadcastJoinThreshold setting is at its default value. Which change causes Spark to choose a broadcast hash join automatically without modifying the query?

• AA. Set spark.sql.autoBroadcastJoinThreshold to 32MB
• BB. Set spark.driver.memory to 64g
• CC. Set spark.sql.autoBroadcastJoinThreshold to 64MB
• DD. Set spark.sql.shuffle.partitions to 64

The bronze table main.bronze.orders has a column items of type ARRAY<STRUCT<sku: STRING, qty: INT>>. The table contains five orders: 1001: 2 items 1002: 1 item 1003: empty array 1004: NULL 1005: 2 items A data engineer runs the following query to build the silver table: SELECT order_id, explode(items) AS item FROM main.bronze.orders How many rows does the query return?

• AA. 4
• BB. 5
• CC. 6
• DD. 7

A data engineer runs the following statements in a notebook attached to a Unity Catalog-enabled workspace. USE CATALOG prod; USE SCHEMA bronze; SELECT * FROM silver.customers; The catalogs dev and prod each contain the schemas bronze and silver, and each of those schemas contains a table named customers. Which table does the query read?

• AA. hive_metastore.silver.customers
• BB. dev.silver.customers
• CC. prod.silver.customers
• DD. prod.bronze.customers

The group analysts has been granted SELECT on the table main.sales.orders. When a member of the group queries the table, the query fails with a permission error. The group holds no other privileges on main or main.sales. What is the minimum additional access that must be granted?

• AA. USE CATALOG on main and USE SCHEMA on main.sales, both to the group
• BB. CREATE TABLE on main.sales to the group, in addition to the existing SELECT
• CC. MODIFY on main.sales.orders to the group, in addition to the existing SELECT
• DD. USE SCHEMA on main.sales to the group, with no change at the catalog level

A DataFrame named df has columns id, code and amt and contains the following six rows. (1, 'a', 10) (1, 'a', 20) (2, 'a', 30) (2, 'b', 30) (3, 'c', 40) (3, 'c', 40) What does df.dropDuplicates(["id", "code"]).count() return?

• AA. 3
• BB. 4
• CC. 5
• DD. 6

Which statement correctly describes the role of Unity Catalog in the Databricks lakehouse?

• AA. It schedules and orchestrates tasks across notebooks, SQL queries and pipelines.
• BB. It centralizes access control, lineage and auditing for data and AI assets across workspaces.
• CC. It stores table data in Parquet files and adds a transaction log that provides ACID guarantees.
• DD. It incrementally ingests new files from cloud storage by using checkpointed state.

A bundle defines a job named nightly_etl with a quartz cron schedule that is not paused. The databricks.yml file contains the following targets. targets: dev: mode: development default: true prod: mode: production A developer runs databricks bundle deploy -t dev. What is the result for the job?

• AA. The job is deployed with its schedule deleted, so it can be started only through the CLI.
• BB. The job is deployed with a [dev <username>] name prefix and its schedule is paused.
• CC. The deployment fails because schedules are permitted only in targets that use mode: production.
• DD. The job is deployed under its original name and runs on its cron schedule in the dev workspace.