Databricks Certification Practice Test 06
Practice Test 06
Databricks Certification Practice Test 06
Practice Test 06
Insightful Saga — Modern Data Engineering Certification Preparation
Question 91
## Question 91 A Delta table already contains data. A Data Engineer attempts to write a DataFrame containing an additional column that does not exist in the destination table. To allow the write operation without manually recreating the table, the engineer should enable **_____**. <Select> <option>Schema Enforcement</option> <option>Schema Evolution</option> <option>Checkpointing</option> <option>Z-Ordering</option> </Select>
Schema Evolution allows compatible schema changes such as adding new columns during write operations. ---
Question 92
## Question 92 A Spark application repeatedly performs a large aggregation that causes memory pressure on executors. Instead of storing the DataFrame only in memory, the engineer chooses to store some partitions on disk when memory becomes full. This is achieved using **_____**. <Select> <option>cache()</option> <option>persist()</option> <option>broadcast()</option> <option>coalesce()</option> </Select>
persist() supports different storage levels and can spill data to disk when memory is insufficient. ---
Question 93
## Question 93 A data platform team wants queries filtering by customer_id to skip reading unnecessary files within the same partition. The team should implement **_____**. <Select> <option>Z-Ordering</option> <option>Checkpointing</option> <option>Cataloging</option> <option>Versioning</option> </Select>
Z-Ordering colocates related values to reduce file scanning during queries. ---
Question 94
## Question 94 A Data Engineer wants a transformation to execute only when an action such as count() or collect() is triggered. This Spark behavior is known as **_____**. <Select> <option>Lazy Evaluation</option> <option>Eager Execution</option> <option>Partition Pruning</option> <option>Broadcasting</option> </Select>
Spark builds the execution plan first and executes only when an action is requested. ---
Question 95
## Question 95 A streaming pipeline performs aggregations over incoming events. To maintain aggregation state across micro-batches, Spark internally maintains **_____**. <Select> <option>Workflow Metadata</option> <option>State Store</option> <option>Driver Cache</option> <option>Cluster Policy</option> </Select>
The State Store tracks aggregation and streaming state between batches. ---
Question 96
## Question 96 A Data Engineer wants to reduce the number of partitions from 500 to 50 without triggering a full shuffle operation. The preferred operation is **_____**. <Select> <option>repartition()</option> <option>coalesce()</option> <option>explode()</option> <option>collect()</option> </Select>
coalesce() is often more efficient when reducing partitions because it minimizes shuffling. ---
Question 97
## Question 97 A company wants to create reusable data transformation pipelines using declarative definitions instead of manually writing orchestration logic. The Databricks solution designed for this is **_____**. <Select> <option>Delta Live Tables</option> <option>Photon</option> <option>Repos</option> <option>SQL Warehouse</option> </Select>
DLT simplifies ETL development, quality enforcement, and pipeline management. ---
Question 98
## Question 98 A Delta table receives updates from multiple users simultaneously. Consistency is maintained through **_____**. <Select> <option>ACID Transactions</option> <option>Executor Caching</option> <option>Workflow Scheduling</option> <option>Notebook Versioning</option> </Select>
ACID guarantees reliable concurrent reads and writes. ---
Question 99
## Question 99 An engineer wants Spark to automatically infer the best number of shuffle partitions during query execution. This optimization is largely supported by **_____**. <Select> <option>Adaptive Query Execution</option> <option>Auto Loader</option> <option>Delta Sharing</option> <option>MLflow</option> </Select>
AQE dynamically adjusts execution strategies while the query runs. ---
Question 100
## Question 100 A notebook contains several transformations but no action. As a result, Spark performs **_____**. <Select> <option>No Execution</option> <option>Full Execution</option> <option>Driver Restart</option> <option>Auto Optimization</option> </Select>
Transformations alone do not trigger execution in Spark. ---
Question 101
## Question 101 A Data Engineer wants to split an array column into multiple rows. The most appropriate Spark function is **_____**. <Select> <option>flatten()</option> <option>explode()</option> <option>collect_set()</option> <option>substring()</option> </Select>
explode() converts array elements into separate rows. ---
Question 102
## Question 102 A query requires ranking employees by salary within each department. The engineer should use a **_____** function. <Select> <option>Aggregate</option> <option>Window</option> <option>Scalar</option> <option>String</option> </Select>
Window functions perform calculations across related rows while retaining row-level detail. ---
Question 103
## Question 103 A team wants pipeline execution to stop automatically whenever mandatory columns contain null values. This requirement is commonly implemented through **_____** rules. <Select> <option>Data Quality</option> <option>Cluster Policy</option> <option>Autoscaling</option> <option>Partitioning</option> </Select>
Data quality rules prevent bad data from reaching downstream consumers. ---
Question 104
## Question 104 A Spark job contains significantly more tasks than available CPU cores. The extra tasks are placed into a **_____** until resources become available. <Select> <option>Queue</option> <option>Schema</option> <option>Catalog</option> <option>Checkpoint</option> </Select>
Spark schedules pending tasks until executor resources become free. ---
Question 105
## Question 105 A company wants machine learning experiments, parameters and metrics to be tracked centrally. Databricks provides **_____** for this purpose. <Select> <option>MLflow</option> <option>Photon</option> <option>Delta Engine</option> <option>State Store</option> </Select>
MLflow manages experiment tracking and model lifecycle management. ---
Question 106
## Question 106 An engineer needs to transform a nested struct column into separate top-level columns. This process is commonly known as **_____** the data. <Select> <option>Flattening</option> <option>Vacuuming</option> <option>Broadcasting</option> <option>Versioning</option> </Select>
Flattening converts complex nested structures into a relational format. ---
Question 107
## Question 107 A Spark application spends excessive time serializing objects between executors. This activity contributes to execution **_____**. <Select> <option>Overhead</option> <option>Governance</option> <option>Lineage</option> <option>Partitioning</option> </Select>
Serialization overhead can significantly impact large-scale Spark workloads. ---
Question 108
## Question 108 A data scientist needs a read-only SQL endpoint dedicated to BI dashboards. The most suitable Databricks compute option is a **_____**. <Select> <option>SQL Warehouse</option> <option>Job Cluster</option> <option>Interactive Notebook</option> <option>Workflow Task</option> </Select>
SQL Warehouses are optimized for BI and analytics workloads. ---
Question 109
## Question 109 A company wants notebook changes to be stored in Git and reviewed through pull requests. Databricks supports this workflow using **_____**. <Select> <option>Repos</option> <option>DLT</option> <option>AQE</option> <option>Photon</option> </Select>
Repos integrates Databricks development with Git providers. ---
Question 110
## Question 110 A large table is being processed by Spark. The engineer wants records with the same key to be redistributed evenly across executors before a transformation. The engineer should use **_____**. <Select> <option>repartition()</option> <option>display()</option> <option>collect()</option> <option>cache()</option> </Select>
repartition() performs a shuffle and redistributes data across partitions more evenly. --- --- title: Databricks Certification Practice Test 07 description: Advanced Multi-Select Questions ---