Free Databricks Databricks-Certified-Data-Engineer-Associate Exam Practice Questions & Explanations

Databricks documents that Delta Sharing does not require data replication, but cloud providers can still charge data egress fees when data is shared across clouds or across geographic regions. Databricks specifically states that sharing within the same region incurs no egress cost, while cross-cloud or cross-region transfers can create additional charges from the underlying cloud provider. That means option A is correct. Options B and C describe same-cloud or same-region use cases that do not trigger the cross-region/cross-cloud egress pattern Databricks calls out. Option D is unrelated to the main documented billing driver for Delta Sharing costs. The important principle is that Delta Sharing itself is designed to avoid replication overhead, but the physical movement of data between cloud boundaries or regions can still result in vendor networking charges. Therefore, when sharing data with external partners in other regions or clouds, engineers should plan for possible egress costs and monitor them accordingly.

Deploying a bundle creates or updates the resources defined in its configuration, but deployment does not automatically execute those resources. The databricks bundle run command starts a deployed job or pipeline by using its resource key from the bundle configuration. Because the job is keyed as etl_job, the correct command is databricks bundle run etl_job. The command bundle deploy run is not valid CLI syntax, and bundle execute is not the command used to run a bundle resource. databricks jobs submit creates and runs a one-time job from a supplied JSON definition rather than executing the job already managed by the bundle. Therefore, option D correctly triggers the deployed job while preserving the bundle-managed resource configuration.

A SQL Serverless Warehouse is the most suitable option for running SQL-based exploratory data analysis at scale while minimizing operational and cost overhead. Serverless SQL warehouses are fully managed by Databricks, meaning Databricks automatically provisions and scales compute resources, applies Photon acceleration, and optimizes query execution without requiring the user to manage infrastructure. This makes them ideal for workloads involving large numbers of ad hoc or interactive SQL queries, which is common during EDA. Because the data resides in a custom-defined cloud network, serverless SQL warehouses can still securely access governed data when Unity Catalog and networking requirements are satisfied. All-purpose clusters (option A) are designed for collaborative development and mixed workloads, not cost-efficient high-concurrency SQL querying. Classic SQL warehouses (option D) are legacy and lack the performance and cost optimizations of newer serverless offerings. Pro SQL warehouses (option B) require user-managed sizing decisions, whereas serverless warehouses optimize both performance and cost automatically. Databricks documentation recommends SQL Serverless Warehouses for fast, scalable, and cost-effective SQL analytics.

The STREAM function is used to process data from a streaming live table or view, which is a table or view that contains data that has been added only since the last pipeline update. Streaming live tables and views are stateful, meaning that they retain the state of the previous pipeline run and only process new data based on the current query. This is useful for incremental processing of streaming or batch data sources. The customers table in the query is a streaming live table, which means that it contains the latest data from the source. The STREAM function enables the query to read the data from the customers table incrementally and create another streaming live table named loyal_customers, which contains the customer IDs of the customers with high loyalty level.Reference:Difference between LIVE TABLE and STREAMING LIVE TABLE,CREATE STREAMING TABLE,Load data using streaming tables in Databricks SQL.

The MERGE INTO command is used to perform upserts, which are a combination of insertions and updates, based on a source table into a target Delta table1. The MERGE INTO command can handle scenarios where the target table cannot contain duplicate records, such as when there is a primary key or a unique constraint on the target table. The MERGE INTO command can match the source and target rows based on a merge condition and perform different actions depending on whether the rows are matched or not.For example, the MERGE INTO command can update the existing target rows with the new source values, insert the new source rows that do not exist in the target table, or delete the target rows that do not exist in the source table1.

The INSERT INTO command is used to append new rows to an existing table or create a new table from a query result2. The INSERT INTO command does not perform any updates or deletions on the existing target table rows.The INSERT INTO command can handle scenarios where the location of the data needs to be changed, such as when the data needs to be moved from one table to another, or when the data needs to be partitioned by a certain column2.The INSERT INTO command can also handle scenarios where the target table is an external table, such as when the data is stored in an external storage system like Amazon S3 or Azure Blob Storage3.The INSERT INTO command can also handle scenarios where the source table can be deleted, such as when the source table is a temporary table or a view4.The INSERT INTO command can also handle scenarios where the source is not a Delta table, such as when the source is a Parquet, CSV, JSON, or Avro file5.

:

1:MERGE INTO | Databricks on AWS

2: [INSERT INTO | Databricks on AWS]

3: [External tables | Databricks on AWS]

4: [Temporary views | Databricks on AWS]

5: [Data sources | Databricks on AWS]