Luyện thi Azure Data Fundamentals DP-900

Structured data fits neatly into tables with defined rows, columns, and types. Semi-structured data such as JSON has flexibility, and unstructured data includes media files.

Semi-structured data has some organization, such as JSON fields or XML tags, but does not require a fixed table schema. Relational data is structured, and media files are unstructured.

Unstructured data, such as images, videos, documents, and free text, has no predefined model. Relational tables and key-value stores impose structure.

A database stores related data in an organized way so applications can insert, update, query, and manage it. Firewalls, servers, and languages are not databases.

OLTP handles many small, fast transactions such as orders and payments. OLAP supports analytical queries over large historical datasets, and media storage or backups are different.

OLAP systems support business intelligence and analytics over large datasets. OLTP systems process short transactions, while printing and email are unrelated.

A data lake stores large volumes of raw data in native formats, including structured, semi-structured, and unstructured data. Relational databases require schema, and spreadsheets or tapes are limited.

A data warehouse consolidates transformed data for reporting and analytics. Queues, networks, and file shares have different purposes.

Batch processing handles data in scheduled chunks, such as nightly jobs. Stream processing handles events as they arrive, and transaction logging records changes.

Stream processing reacts to events in near real time. Batch processing waits for scheduled chunks, and indexing or cold storage are storage concepts.

Tables are the core relational structure, with rows as records and columns as attributes. JSON documents, graph edges, and storage accounts are non-relational concepts.

A primary key uniquely identifies rows so no duplicate record exists. Large columns, backups, and addresses do not provide relational identity.

A foreign key references a primary key in another table to create relationships. It does not store passwords, duplicate rows, or backups.

SQL, or Structured Query Language, is used for relational database queries and management. HTML and CSS describe web pages, and YAML is a configuration format.

Azure SQL Database is a managed relational database with automated backups, patching, and high availability. Blob and Queue Storage are non-relational, and Front Door routes traffic.

Azure Database for MySQL and Azure Database for PostgreSQL provide managed open-source relational databases. Cosmos DB, Table Storage, Files, and Blob Storage are non-relational or file services.

Running SQL Server on Azure VMs is IaaS because you manage the operating system and database software. PaaS manages the platform, SaaS delivers software, and FaaS runs functions.

ACID properties ensure reliable transactions: atomicity, consistency, isolation, and durability. The other phrases are not the ACID acronym.

SELECT retrieves rows and columns from tables. INSERT adds data, DELETE removes data, and CREATE defines objects.

Primary keys identify rows, foreign keys create links, tables organize data, and SQL queries relational data.

Document databases store JSON-like documents with flexible fields. Relational tables and SQL views require fixed schemas.

Key-value stores map unique keys to values, offering fast lookups. Documents, graphs, and relational tables have different structures.

Graph databases store nodes and edges to model relationships such as social networks. Documents, key-value pairs, and column families have different shapes.

Columnar NoSQL databases group data by columns for large analytical workloads. Relational, graph, and document models organize data differently.

Azure Cosmos DB supports multiple data models including document, key-value, graph, and columnar. SQL Database and PostgreSQL are relational, and Synapse is analytics.

Blob Storage stores unstructured objects at scale. SQL Database is relational, Files provides shares, and Queue Storage stores messages.

Queue Storage holds messages that producers and consumers process asynchronously. Table Storage is NoSQL, Files provides shares, and Cosmos DB is a database.

Azure Files provides managed SMB and NFS file shares. Blob is object storage, Queue is messaging, and Table is NoSQL key-value storage.

Table Storage is a schemaless NoSQL key-value store. SQL Database is relational, Synapse is analytics, and Data Factory orchestrates pipelines.

Documents store flexible records, key-value stores map keys to values, graphs store nodes and edges, and columnar stores use wide columns.

Azure Synapse Analytics unifies SQL warehouses, Apache Spark, pipelines, and visualization. App Service hosts apps, DNS resolves names, and Bastion secures VM access.

Data Factory creates and schedules pipelines that copy and transform data. Blob Storage stores data, VMs run workloads, and Policy enforces rules.

Stream Analytics processes streaming events with SQL-like queries. Backup protects data, Traffic Manager routes traffic, and DNS resolves names.

Event Hubs ingests and buffers millions of streaming events for downstream processing. Files, Table Storage, and CDN serve other purposes.

Azure Databricks offers a fast, collaborative Apache Spark platform for data engineering and analytics. App Service hosts web apps, Front Door routes traffic, and Monitor observes systems.

ELT extracts data, loads it into the target system, and transforms it there. The other combinations are not the standard ELT meaning.

ETL extracts data, transforms it before loading, and then loads it into the destination. The other acronyms are incorrect.

Azure Data Lake Storage provides hierarchical, scalable storage built for analytics. Queue and Table Storage serve messaging and NoSQL needs, and CDN caches content.

A lakehouse stores raw data like a lake while adding warehouse features such as schema management and SQL analytics. It is not a room, view, or queue.

Data Factory orchestrates pipelines, Stream Analytics queries streams, Synapse provides analytics, and Event Hubs ingests events.

Power BI is the Microsoft business intelligence tool for reports and dashboards. DevOps, VS Code, and Terminal are developer tools.

Power BI Desktop is the authoring environment for reports. Power BI Service hosts and shares reports, and the other tools manage Azure or SQL Server.

Power BI Service is the cloud workspace where reports, dashboards, and datasets are shared. Desktop is for authoring, and Explorer or Blob Storage do not host Power BI reports.

A dataset contains the data and relationships Power BI uses in reports. Security rules, templates, and VMs are unrelated.

A dashboard is a single page that displays key visualizations from one or more reports. Queries, sources, and backups are not dashboards.

Measures use Data Analysis Expressions (DAX) to calculate values such as totals and percentages dynamically. Labels, tables, and ports are not measures.

Row-level security filters data by user identity so each user sees only permitted rows. It does not block all access or alter report text.

Scheduled refresh pulls updated data from sources automatically. Printing, themes, and web publishing do not refresh data.

Workspaces organize and share Power BI content with teams. They are not disks, DNS zones, or keys.

Desktop authors reports, Service shares them, measures calculate values, and dashboards display visuals on one page.

Microsoft Purview provides data cataloging, classification, lineage, and governance. DevOps is for software delivery, Defender is security, and CDN is content delivery.

Lineage shows data origins and transformations across pipelines. Deduplication, encryption, and backups are separate processes.

Classification assigns sensitivity labels such as public, internal, or confidential. Sorting, compression, and type changes are technical operations.

Dynamic data masking limits sensitive data exposure in query results while preserving the stored value. Row-level security filters rows, and scaling or balancing are availability features.

Encryption at rest protects stored data using service or customer-managed keys. TLS protects data in transit, and DNSSEC or SMB signing serve other purposes.

TLS encrypts data in transit between clients and services. Encryption at rest protects stored data, BitLocker encrypts disks, and masking hides values.

RBAC assigns roles such as Reader, Contributor, and Owner to control resource access. NSGs filter traffic, Front Door routes traffic, and Monitor observes systems.

Microsoft Entra ID provides authentication and authorization. DNS resolves names, Blob stores data, and VNets provide networking.

Data residency ensures data is stored in approved regions for compliance or performance. Deletion, compression, and access grants are different.

Purview catalogs data, lineage tracks changes, masking hides sensitive values, and RBAC controls access.