Microsoft Certified: Azure Data Fundamentals practice questions

Question 1

A streaming squad is tuning the hot path of a Lambda architecture and must decide what trade-off the speed layer accepts in order to deliver real-time results. Per Microsoft's big data architecture guidance, which statement correctly describes the speed layer?

Accepted Answer

The speed layer analyzes data in real time and is designed for low latency, at the expense of some accuracy in its results.

Answer

The speed layer guarantees the same accuracy as the batch layer because it fully reprocesses the entire historical dataset on every event.

Answer

The speed layer permanently stores the master copy of all raw data and feeds it directly into long-term cold archival storage for audits.

Answer

The speed layer only indexes the batch view so historical queries return faster, which removes any need for a separate serving layer.

Question 2

A company needs to give executives a single at-a-glance view that consolidates on-premises sales data and cloud data and refreshes its tiles automatically as the underlying data changes. Which Power BI artifact best fits this need?

Accepted Answer

A Power BI dashboard, a single canvas that consolidates tiles from on-premises and cloud sources

Answer

A Power BI report page, a multi-page canvas focused on the deep exploration of one semantic model

Answer

A Power BI dataflow, a reusable transformation that prepares data before it loads into a model

Answer

A Power BI paginated report, a print-ready layout built for pixel-perfect operational documents

Question 3

A startup is migrating a small PostgreSQL development workload to Azure Database for PostgreSQL and is evaluating the Burstable compute tier, but the architect worries about sustained CPU load. What is the documented risk?

Accepted Answer

Burstable uses a CPU credit model, so once credits run out the VM is capped at baseline CPU and sustained load degrades.

Answer

Burstable reserves full vCores continuously, so sustained load always runs at peak speed but roughly doubles the monthly bill.

Answer

Burstable auto-scales out to read replicas under load, so sustained CPU pressure is absorbed without any throttling.

Answer

Burstable shifts overflow load to a Hyperscale storage tier, so sustained CPU demand never reaches the primary compute.

Question 4

A platform team is designing the cost model for an Azure Cosmos DB workload and must decide how to express the cost of reads, writes, and queries as a single normalized unit. Which option correctly describes how Azure Cosmos DB measures the cost of database operations?

Accepted Answer

Request Units per second (RU/s), a performance currency that normalizes the cost of every read, write, and query operation.

Answer

Database Transaction Units (DTUs), the blended compute-and-storage metric that Azure SQL Database uses to size its service tiers.

Answer

Throughput Units, the ingestion-capacity metric that Azure Event Hubs bills for streaming event pipelines per namespace.

Answer

Provisioned vCores, the dedicated processor count that Azure SQL Managed Instance reserves for each compute generation.

Question 5

A company is evaluating where Azure SQL Managed Instance fits between PaaS Azure SQL Database and an Azure VM running SQL Server. Per the Microsoft Learn module on data services, which option correctly describes Azure SQL Managed Instance?

Accepted Answer

Azure SQL Managed Instance – a hosted instance of SQL Server with automated maintenance, which allows more flexible configuration than Azure SQL DB but with more administrative responsibility for the owner.

Answer

Azure SQL Managed Instance – a fully unmanaged offering that requires the customer to install, patch, and back up SQL Server on their own hardware at all times.

Answer

Azure SQL Managed Instance – a NoSQL key-value store used to back high-throughput shopping cart sessions and other ephemeral data.

Answer

Azure SQL Managed Instance – an offering identical to Azure SQL Database in every configurable dimension, including admin surface and feature parity with SQL Server.

Question 6

A retail analytics team is being asked to characterize Azure Synapse Analytics at the platform level before standing up an enterprise warehouse for sales data. Per the Microsoft Learn overview of Azure Synapse Analytics, which option correctly describes the service?

Accepted Answer

Azure Synapse is an enterprise analytics service that accelerates time to insight across data warehouses and big data systems.

Answer

Azure Synapse is a strictly OLTP key-value store designed for low-latency point lookups in line-of-business workloads and provides no Spark, SQL warehouse, or pipeline capabilities for analytical processing.

Answer

Azure Synapse is a virtual network configuration service used to peer VNets across Azure regions and is unrelated to data warehouses or big data processing on the Microsoft cloud platform.

Answer

Azure Synapse is a Kubernetes orchestration service used to run containerized workloads and does not include any SQL, Spark, or pipeline capabilities for analytics on Azure.

Question 7

A SaaS company is rolling out hundreds of small tenant databases on Azure SQL Database whose usage spikes are unpredictable and rarely peak at the same time, and the team must control total spend. Which Azure SQL Database capability best fits?

Accepted Answer

Place the tenant databases in an elastic pool so they share a set resource allotment at a set price.

Answer

Provision each tenant on its own Hyperscale database so storage scales out independently as the data grows.

Answer

Put each tenant on a separately purchased single vCore database and right-size each one as demand shifts.

Answer

Use active geo-replication to spread the bursty tenant read load across paired regional secondaries.

Question 8

An organization is designing an event-driven pipeline that must react to every insert and update in an Azure Cosmos DB container in the order the changes occurred. Per the Microsoft Learn documentation on the change feed, which option correctly describes it?

Accepted Answer

A persistent record of container changes in order, which consumers can process asynchronously and incrementally.

Answer

A nightly batch job that exports the full container to Azure Blob Storage so downstream systems can reload it later.

Answer

A trigger that automatically rolls back any write violating a unique key constraint before it is committed.

Answer

A read-only replica kept in a secondary region that is used only for disaster-recovery failover events.

Question 9

An organization is migrating an on-premises enterprise data warehouse to Azure Synapse Analytics and the architect must size the compute for a dedicated SQL pool. Per Microsoft's Synapse data warehousing guidance, which statement is correct?

Accepted Answer

The size of a dedicated SQL pool is set in Data Warehousing Units (DWU), and it stores data in relational tables using columnar storage.

Answer

A dedicated SQL pool is billed strictly per query scanned and stores all of its data as JSON documents to support schema-on-read analytics.

Answer

A dedicated SQL pool is sized in Request Units (RU/s) and uses row-based storage that is tuned for single-record OLTP write workloads.

Answer

A dedicated SQL pool autoscales fully transparently and persists its data only in Azure Blob Storage instead of in any relational tables.

Question 10

A data analyst is exploring Power BI for the first time and is being asked what a visualization actually is in a Power BI report. Per the Microsoft Learn overview of visualizations in Power BI, which option correctly describes a visual in Power BI?

Accepted Answer

Visualizations, also called visuals, are the building blocks of Power BI reports, and they transform the curated data and DAX calculations from your Power BI semantic model into meaningful insights.

Answer

Visualizations are an obsolete component of Power BI that has been removed from all reports since Microsoft Fabric became generally available, and they no longer exist in Power BI Desktop.

Answer

Visualizations are a Cosmos DB for Gremlin feature used to render social-network graphs and are unrelated to Power BI reports or semantic models.

Answer

Visualizations are a Kubernetes-only feature used to render real-time pod metrics and are not part of the Power BI report authoring experience.

Question 11

A company is evaluating the DTU-based purchasing model for a new Azure SQL Database and an architect must brief stakeholders on what a single DTU actually measures. Per the Microsoft Learn DTU-based purchasing model documentation, which option correctly describes a database transaction unit?

Accepted Answer

A blended measure of CPU, memory, and read and write operations on the database.

Answer

A unit that meters only the storage that the single database instance has consumed.

Answer

A measure of the network egress bandwidth that is allocated to the database per second.

Answer

A count of the concurrent user connections that the database service tier permits at once.

Question 12

A company is migrating an existing MongoDB-backed application to Azure with minimal code changes and the architect is being asked which Cosmos DB API is the best fit. Per the Microsoft Learn introduction to Azure Cosmos DB for MongoDB, which option correctly describes the API?

Accepted Answer

Azure Cosmos DB for MongoDB is a fully managed NoSQL database service that enables you to run MongoDB workloads in the cloud with minimal changes to your application code, supporting the MongoDB wire protocol.

Answer

Azure Cosmos DB for MongoDB is a SQL Server-compatible relational engine that requires the application to be rewritten in T-SQL before it can be used on Azure.

Answer

Azure Cosmos DB for MongoDB is a Cassandra-only API that does not support the MongoDB wire protocol and prohibits the use of existing MongoDB drivers.

Answer

Azure Cosmos DB for MongoDB is a graph traversal engine that exposes only the Gremlin query language and is unrelated to MongoDB workloads or drivers.

Question 13

A junior data engineer is designing an ingestion layer for a manufacturing telemetry pipeline and the team is being asked which delimited file format Microsoft Learn lists as the most common for delimited data. Per the data fundamentals module on file storage, which option matches?

Accepted Answer

The most common format for delimited data is comma-separated values (CSV) in which fields are separated by commas, and rows are terminated by a carriage return / new line, with the first line optionally containing field names.

Answer

The most common format for delimited data is Azure Cosmos DB Change Feed JSON, which is emitted automatically whenever any container is modified and is the default ingestion format for Azure Data Factory.

Answer

The most common format for delimited data is Apache Parquet, a columnar binary format that includes a schema descriptor at the end of every file regardless of whether the data is structured or semi-structured.

Answer

The most common format for delimited data is the Microsoft Excel binary workbook (XLSB) format, which Microsoft recommends for all enterprise-scale data interchange between Azure services.

Question 14

An administrator is being asked to ensure that each regional manager who opens a shared Power BI report can see only the rows for their own region, without maintaining a separate copy of the report per region. Which Power BI feature accomplishes this?

Accepted Answer

Row-level security, which restricts the data a given user can access inside one shared report

Answer

Sensitivity labels, which classify and tag a report's data for compliance and protection reporting

Answer

A model relationship, which propagates filters from one related table to another in the model

Answer

Always Encrypted, which protects data in use through client-side encryption of stored columns

Question 15

A platform team is evaluating the vCore-based purchasing model for a database that runs a steady, predictable load around the clock and must choose between the provisioned and serverless compute tiers on cost. Per the Microsoft Learn purchasing models documentation, which choice is correct?

Accepted Answer

Provisioned, because its vCore unit price per unit of time is lower than the serverless tier's.

Answer

Serverless, because its per-second billing is always cheaper than provisioned for any workload.

Answer

Either tier costs exactly the same per vCore, so the decision has no impact on the monthly bill.

Answer

Serverless, because a constantly busy database benefits most from automatic pausing during idle time.

Microsoft Certified: Azure Data Fundamentals

Sample questions

Sources

Similar exams