DP-900 sample questions with answers
The other 173 questions are in the practice tool, with answers revealed as you go.
Start 20 questionsDescribe core data concepts
- A media company is cataloguing its data assets. Which asset below counts as unstructured data?
- Which description correctly contrasts schema-on-write with schema-on-read?
- A travel marketplace stores hotel listings whose attributes differ by property type and change often as partners add new fields. The catalog must be readable with low latency on several continents and stay available if one region fails. Which data type and store combination fits?
- A payments processor must guarantee that refunds and charges are recorded exactly once, support auditors who run cross-customer analysis over several years, and process card authorizations in real time. Which architecture best meets these needs?
- Which role is accountable for keeping a company's production SQL Server instances available, including patching, access security, backup and restore testing, and performance tuning?
- A payroll export tool saves employee records as a plain-text file in which each line is one record and commas divide the fields. Which file format is this?
Relational data on Azure
- A healthcare startup has to run a SQL Server instance inside a virtual machine in Azure. It also needs administrator rights on the guest Windows operating system so it can install a third-party monitoring agent. Which option meets these requirements?
- An order-line table stores a product code that must always match a code that exists in the product catalog table. Which database construct enforces this link between the two tables?
- An HR team wants staff to query employee records without seeing salary columns. The administrator defines a saved SELECT that exposes only the permitted columns. Which object does this describe?
- An Azure SQL Database must keep serving requests if an entire datacenter fails within one region, and the team does not want to build any replicas by hand. Which configuration provides this?
- A web application opens and closes thousands of short database sessions every minute against an Azure Database for PostgreSQL flexible server, and the server keeps hitting its connection limit. Which technique reduces this overhead?
- A telecom keeps a customers table with 400 million rows. Plan A splits the table so customers from different regions sit in separate partitions. Plan B moves rarely used, very wide attachment columns into a companion table keyed by the same customer ID. Which statement correctly names these two approaches?
- A manufacturer wants a managed SQL database running on its own Kubernetes cluster inside a factory, because the data must stay on site. The company still wants to view and govern that instance from the Azure portal. Which offering matches this need?
- Sales representatives in every region query one shared orders table. Each representative should see only orders from their own territory, and the filtering must happen inside the database so no application code changes. Which feature should be applied?
Non-relational data on Azure
- A team runs a product catalog built on existing MongoDB drivers and wants to move it onto a managed Azure service while keeping its application code and query syntax largely unchanged. Which Cosmos DB API should it choose?
- A banking app needs reads to trail writes by no more than a fixed number of updates or a set time window, while performing better than strong consistency. Which Cosmos DB consistency level matches?
- A utility collects smart meter readings in a flexible schema, and some meters add extra fields over time. Lookups need to be fast by meter ID and reading time. Which option is most appropriate?
- A media archive stores uploaded videos, PDFs, and JPEG photos in Azure Storage. Each file is uploaded whole and read back whole, and the archive never appends to files or backs a virtual machine disk. Which blob type should it use?
Analytics workloads on Azure
- A data engineering group needs a managed environment for transforming terabytes of event logs and training models, built around an open-source distributed engine. Which Azure service matches?
- Microsoft folded a massively parallel warehouse service into Synapse Analytics as a dedicated pool. What was that service called before it was renamed?
- A company using Microsoft Fabric wants its warehouses, lakehouses, and reports to read the same underlying files without copying data between silos. Which component provides that shared storage layer?
- A machine learning team needs a shared, notebook-based environment where it can train models on very large tables using Python and Scala, then share its work with colleagues. Which Azure service should it use?
- An Azure Data Factory author needs two kinds of data preparation: a visual transformation graph that runs at scale on Spark, and an interactive, spreadsheet-style cleanup experience for a business user. Which statement correctly describes these two options?
- An online retailer stores orders in Azure Cosmos DB and wants near real-time sales dashboards in Azure Synapse Analytics. The checkout workload must not slow down. Which benefit does Azure Synapse Link provide?
- A payments company must score card transactions for fraud in real time. Requirements: ingest millions of events per second, run a complex ML model on each event, send immediate alerts while also keeping history for later analysis, and guarantee that each transaction is recorded exactly once. Which architecture should it implement?
- A company must copy nightly sales extracts from an on-premises SQL Server and several SaaS APIs into a data lake, then run a transformation step only after each copy finishes. Which Azure service coordinates these steps?
- A research group must analyze several terabytes of sensor logs, which is far too large for one relational server to handle efficiently. They plan to spread the work across a cluster of machines using a distributed engine. What is this approach called?