SESION 9 - Examen de Práctica de Databricks

Which of the following locations houses the controller and worker nodes of a cluster managed by Databricks?

• Data plane

• Control plane

• Databricks file system

• JDBC data source

• Databricks web application

Two junior data engineers are drafting separate parts of a single data pipeline notebook. They are working on separate Git branches to be able to pair program in the same notebook simultaneously. A senior data engineer with experience in Databricks suggests that there is a better alternative for this type of collaboration. Which of the following supports the senior data engineer's statement?

• Databricks notebooks support automatic change tracking and versioning.

• Databricks notebooks support real-time co-authoring in a single notebook.

• Databricks notebooks support comments and notification comments.

• Databricks notebooks support the use of multiple languages in the same notebook.

• Databricks notebooks support the creation of interactive data visualizations.

Which of the following describes how Databricks Repos can help facilitate CI/CD workflows on the Databricks Lakehouse Platform?

• Databricks Repos can facilitate the pull request, review, and approval process before merging branches.

• Databricks Repos can merge changes from a Git secondary branch into a Git main branch.

• Databricks Repos can be used to design, develop, and activate Git automation pipelines.

• Databricks Repos can store the Git repository as the single source of truth.

• Databricks Repos can commit or push code changes to trigger a CI/CD process.

A data architect has determined that a table in the following format is necessary:

What of the following code blocks uses SQL DDL commands to create an empty Delta table in the above format, regardless of whether a table with this name already exists?

• CREATE OR REPLACE TABLE table_name AS SELECT id STRING, birthDate DATE, avgRating FLOAT USING DELTA

• CREATE OR REPLACE TABLE table_name (id STRING, birthDate DATE, avgRating FLOAT)

• CREATE TABLE IF NOT EXISTS table_name (id STRING, birthDate DATE, avgRating FLOAT)

• CREATE TABLE table_name AS SELECT id STRING, birthDate DATE, avgRating FLOAT

• CREATE OR REPLACE TABLE table_name WITH COLUMNS (id STRING, birthDate DATE, avgRating FLOAT) USING DELTA

A data engineering team needs to query a Delta table to extract rows that all meet the same condition. However, the team has noticed that the query is running slowly. The team has already adjusted the size of the data files. After investigating, the team concluded that the rows meeting the condition are located sparsely throughout each of the data files. Based on the scenario, which of the following optimization techniques could speed up the query?

• Data skipping

• Z-Ordering

• Bin packing

• Write as a Parquet file

• Tuning the file size

A data engineer wants to create a relational object by extracting data from two tables. The relational object should be used by other data engineers in other sessions. To save on storage costs, the data engineer wants to avoid copying and storing physical data. Which of the following relational objects should the data engineer create?

• View

• Temporary view

• Delta table

• Database

• Spark SQL table

A data engineer has ingested a JSON file into a raw_table with the following schema:

cart_id STRING,

items ARRAY <item_id:STRING>

The junior data engineer would like to unnest the items column in raw_table to result in a new table with the following schema:

cart_id STRING, item_id STRING.

Which of the following commands should the junior data engineer run to complete this task?

• SELECT cart_id, filter(items) AS item_id FROM raw_table;

• SELECT cart_id, flatten(items) AS item_id FROM raw_table;

• SELECT cart_id, reduce(items) AS item_id FROM raw_table;

• SELECT cart_id, explode(items) AS item_id FROM raw_table;

• SELECT cart_id, slice(items) AS item_id FROM raw_table;

A data engineer has ingested a JSON file into a raw_table with the following schema: transaction_id STRING, payload ARRAY. The data engineer wants to efficiently extract the date of each transaction into a table with the following schema: transaction_id STRING, date TIMESTAMP. Which of the following commands should the data engineer execute to complete this task?

• SELECT transaction_id, explode(payload) FROM raw_table;

• SELECT transaction_id, payload.date FROM raw_table;

• SELECT transaction_id, date FROM raw_table;

• SELECT transaction_id, payload[date] FROM raw_table;

• SELECT transaction_id, date from payload FROM raw_table;

A data analyst has provided the data engineering team with the following Spark SQL query:

SELECT district,

avg(sales) FROM store_sales_20220101

GROUP BY district;

The data analyst would like the data engineering team to run this query every day. The date at the end of the table name (20220101) should be automatically replaced with the current date each time the query is run. Which of the following approaches could be used by the data engineering team to efficiently automate this process?

• They could wrap the query using PySpark and use Python's string variable system to automatically update the table name.

• They could manually replace the date within the table name with the current date of the day.

• They could request the data analyst to rewrite the query to run less frequently.

• They could replace the date formatted as a string in the table with a date formatted as a timestamp.

• They could pass the table to PySpark and develop a robustly tested module on the existing query.

A data engineer needs to dynamically create a table name string using three Python variables: region, store, and year. An example of a table name is below when region = "nyc", store = "100", and year = "2021":
nyc100_sales_2021.
What command should the data engineer use to construct the table name in Python?

• A. "{region}+{store}+_sales_+{year}"

• B. f"{region}+{store}+_sales_+{year}"

• C. "{region}{store}_sales_{year}"

• D. f"{region}{store}_sales_{year}"

• E. {region}+{store}+"_sales_"+{year}