DP-900 · Glossary
The terms, in one sentence each
124 terms from the DP-900 study guide. Each one links to the card that explains it.
A
- access tier also: access tiers
- Hot, Cool, Cold or Archive: how a blob is kept. A cooler tier costs less to store and more, or longer, to read. Study card: Azure Blob storage →
- ACID
- Atomicity, consistency, isolation and durability: the four guarantees that make a transaction trustworthy. Study card: Transactional workloads →
- Apache Spark also: Spark
- An open source engine that processes large volumes of data in parallel across a cluster of machines. Study card: Fabric and Azure Databricks →
- append blob also: append blobs
- The blob type that only accepts new data at the end, which suits log files. Study card: Azure Blob storage →
- attribute also: attributes
- One characteristic of an entity, such as a customer's name or city. In a table it is a column. Study card: Structured data →
- Avro
- A row based file format. The file carries its own schema in a JSON header, followed by compact binary records. Study card: Data file formats →
- Azure Blob storage also: Blob storage
- The Azure service for unstructured data: files of any kind, stored as blobs inside containers. Study card: Azure Blob storage →
- Azure Cosmos DB also: Cosmos DB
- A fully managed, globally distributed nonrelational database with fast reads and writes. Study card: Cosmos DB use cases →
- Azure Data Factory also: Data Factory
- A service for building pipelines that copy and transform data between stores. Study card: Azure datastore options →
- Azure Database for MySQL
- A managed service for the open source MySQL database engine. Study card: Azure open-source databases →
- Azure Database for PostgreSQL
- A managed service for the open source PostgreSQL engine, which can also store custom and geometric data types. Study card: Azure open-source databases →
- Azure Databricks also: Databricks
- An analytics platform based on Apache Spark that runs inside your own Azure subscription. Study card: Fabric and Azure Databricks →
- Azure Event Hubs also: Event Hubs
- A service that takes in very large numbers of events per second, ready for stream processing. Study card: Real-time analytics services →
- Azure File Sync
- A service that keeps a cached copy of a file share on a local server synchronized with Azure Files. Study card: Azure Files →
- Azure Files
- Cloud file shares that several users or applications can mount at the same time, like a file server. Study card: Azure Files →
- Azure IoT Hub also: IoT Hub
- An event ingestion service like Event Hubs, built for connecting and managing IoT devices. Study card: Real-time analytics services →
- Azure SQL Database also: SQL Database
- A fully managed relational database service, built for new cloud applications. It needs the least management of the Azure SQL family. Study card: Azure SQL family →
- Azure SQL Managed Instance also: SQL Managed Instance, Managed Instance
- A managed SQL Server instance with near full compatibility, for moving existing SQL Server systems with few changes. Study card: Azure SQL family →
- Azure Storage also: storage account, storage accounts
- The Azure service that holds blobs, file shares and tables. A storage account is the unit you create and pay for. Study card: Azure datastore options →
- Azure Stream Analytics also: Stream Analytics
- A PaaS service that runs a continuous query over a stream of events. Study card: Real-time analytics services →
- Azure Table storage also: Table storage
- A key-value store in Azure Storage: rows are found by a partition key and a row key, and may have different columns. Study card: Azure Table storage →
B
- batch processing also: batch
- Collecting records first and processing the whole group together, on a schedule or when a threshold is reached. Study card: Batch versus streaming data →
- BLOB also: BLOBs, binary large object, binary large objects
- Binary large object: a file of binary data, such as an image, a video or a backup. Study card: Unstructured data →
- block blob also: block blobs
- The blob type for ordinary files: large binary objects that change rarely. Study card: Azure Blob storage →
- BSON
- A binary form of JSON. It is the document format that MongoDB uses. Study card: Cosmos DB APIs →
C
- column family also: column families
- A nonrelational store of rows whose columns are grouped into related families. Rows do not all need the same columns. Study card: Common data stores →
- columnar
- Stored by column instead of by row, so a query reads only the columns it needs. Study card: Data file formats →
- composite key also: composite keys
- A key made of two or more columns together, used when no single column is unique. Study card: Normalization →
- consistency level also: consistency levels
- A setting that trades how fresh the data you read is against speed and availability, from Strong to Eventual. Study card: Cosmos DB use cases →
- Copilot
- Microsoft's AI assistant. In Power BI it can summarize a report, create report pages and write DAX from a description. Study card: Power BI capabilities →
- Cosmos DB for MongoDB
- The Azure Cosmos DB API that accepts the query language and the drivers of MongoDB. Study card: Cosmos DB APIs →
- Cosmos DB for NoSQL
- The native API of Azure Cosmos DB: JSON documents, queried with a language that looks like SQL. Study card: Cosmos DB APIs →
- CQL
- Cassandra Query Language: the query language of Apache Cassandra, accepted by the Azure Cosmos DB API for Cassandra. Study card: Cosmos DB APIs →
- CRUD
- Create, retrieve, update and delete: the four basic operations on records. Study card: Transactional workloads →
D
- dashboard also: dashboards
- A single page in the Power BI service that brings together visuals from one or more reports. Study card: Power BI capabilities →
- data analyst also: data analysts
- The role that explores data and turns it into models, reports and visualizations that support decisions. Study card: Data analyst role →
- data engineer also: data engineers
- The role that builds and monitors the pipelines that ingest, cleanse and deliver data. Study card: Data engineer role →
- data lake also: data lakes
- File storage that holds data in any format and applies a structure only when the data is read. Study card: Analytical data stores →
- data lakehouse also: lakehouse, lakehouses
- Data lake storage with a SQL endpoint and an enforced schema on top, made possible by the Delta Lake format. Study card: Analytical data stores →
- data type also: data types
- The kind of value a column accepts, such as text, a whole number, a decimal number or a date. Study card: Relational data model →
- data warehouse also: data warehouses, warehouse
- A relational store whose schema is designed for analytics and reporting, typically with fact and dimension tables. Study card: Analytical data stores →
- database administrator also: database administrators, DBA
- The role that keeps databases available, secure, backed up and performing well. Study card: Database administrator role →
- DAX
- Data Analysis Expressions: the formula language for measures and calculated columns in Power BI. Study card: Power BI capabilities →
- DCL also: Data Control Language
- Data Control Language: statements that manage permissions, such as GRANT, DENY and REVOKE. Study card: SQL statement types →
- DDL also: Data Definition Language
- Data Definition Language: statements that create, change or remove database objects, such as CREATE, ALTER and DROP. Study card: SQL statement types →
- decomposition tree
- A Power BI visual that breaks a value down across several dimensions, level by level. Study card: Data visualization types →
- delimited text also: CSV
- A plain text file with one record per line and a separator, often a comma, between the fields. CSV is the best known kind. Study card: Data file formats →
- Delta Lake
- A storage layer that adds a transaction log to Parquet files, so a data lake gets versioning and reliable updates. Study card: Data file formats →
- denormalized also: denormalization
- Keeping related data together in one row, with repetition, instead of splitting it over several tables. Study card: Normalization →
- dimension also: dimensions
- An entity you group and filter measures by, such as product, customer or date. Study card: Power BI data models →
- dimension table also: dimension tables
- A table that describes an entity, with a unique key and the descriptive attributes you group by. Study card: Power BI data models →
- Direct Lake also: Direct Lake mode
- A Power BI storage mode that reads tables in OneLake directly, without importing or refreshing a copy. Study card: Power BI capabilities →
- DML also: Data Manipulation Language
- Data Manipulation Language: statements that read or change rows, such as SELECT, INSERT, UPDATE and DELETE. Study card: SQL statement types →
- document database also: document databases
- A nonrelational database that stores each entity as one JSON document and can query inside the document. Study card: Common data stores →
E
- elastic pool also: elastic pools
- A group of Azure SQL databases that share one set of resources, which suits databases whose load varies. Study card: Azure SQL family →
- ELT
- Extract, load, transform: the data is loaded as it is and reshaped inside the store afterward. Study card: Data ingestion and processing →
- entity also: entities
- A thing you keep data about, such as a customer, a product or an order. Study card: Structured data →
- ETL
- Extract, transform, load: the data is reshaped before it is loaded into the store. Study card: Data ingestion and processing →
- Eventhouse
- The store for event and time series data in Fabric Real-Time Intelligence, queried with KQL. Study card: Real-time analytics services →
F
- fact table also: fact tables
- The table of recorded events, one row per event, that holds the numbers you measure. Study card: Power BI data models →
- Flexible Server
- A deployment option of the managed MySQL and PostgreSQL services that gives more control over configuration and cost. Study card: Azure open-source databases →
- foreign key also: foreign keys
- A column that holds the primary key of a row in another table, which links the two rows. Study card: Normalization →
G
- globally distributed
- The data is replicated to several regions of the world, so users read and write close to where they are. Study card: Cosmos DB use cases →
- graph database also: graph databases
- A nonrelational database of entities, called nodes, and the relationships between them. Study card: Common data stores →
- Gremlin
- A language for traversing graph data, used by the Azure Cosmos DB API for Apache Gremlin. Study card: Cosmos DB APIs →
H
- hierarchical namespace
- A setting on a storage account that organizes blobs in real folders, which turns the account into a data lake. Study card: Azure datastore options →
- hierarchy also: hierarchies
- Levels within a dimension, such as year, month and day, that let a report drill up and down. Study card: Power BI data models →
I
- IaaS also: infrastructure as a service
- Infrastructure as a service: you rent virtual machines and manage the operating system and the software on them yourself. Study card: Azure SQL family →
- index also: indexes
- A sorted lookup structure on a column that lets the database find rows without scanning the whole table. Study card: Database objects →
- ingestion also: ingest, ingests, ingested
- Moving data from its sources into an analytical store, such as a data lake or a data warehouse. Study card: Data ingestion and processing →
J
- JSON
- JavaScript Object Notation: a text format of named fields that can hold nested objects and lists. Study card: Semi-structured data →
K
- key influencers
- A Power BI visual that shows which factors most strongly drive a chosen metric. Study card: Data visualization types →
- key-value also: key value
- A store where every item is a unique key with a value attached. You find an item by its key. Study card: Common data stores →
- KQL also: Kusto Query Language
- Kusto Query Language: a query language for exploring large volumes of log and event data. Study card: Real-time analytics services →
L
- lambda architecture
- A design that processes the same data twice: in real time for live views and in batches for historical analysis. Study card: Batch versus streaming data →
- latency
- The delay between something happening and its result being available. Study card: Batch versus streaming data →
- lifecycle management policy also: lifecycle policy
- A rule that moves blobs to a cheaper access tier, or deletes them, automatically as they age. Study card: Azure Blob storage →
- linked service also: linked services
- A stored connection to a data source or destination, which the activities of a pipeline use. Study card: Data ingestion and processing →
M
- measure also: measures
- A numeric value you analyze, such as revenue or quantity sold, calculated by aggregating rows of a fact table. Study card: Power BI data models →
- medallion architecture also: medallion
- A way to organize a lakehouse in three layers: bronze for raw data, silver for cleansed data and gold for reporting ready data. Study card: Analytical workloads →
- Microsoft Fabric also: Fabric
- Microsoft's analytics platform in one SaaS workspace: ingestion, lakehouse, warehouse, real-time analytics and Power BI on one shared lake. Study card: Fabric and Azure Databricks →
N
- NFS also: Network File System
- Network File System: a file sharing protocol for Linux clients. Study card: Azure Files →
- nonrelational database also: nonrelational databases, nonrelational, NoSQL
- A database that does not use tables with a fixed relational schema. The four common kinds are key-value, document, column family and graph. Study card: Common data stores →
- normalization also: normalized, normalizing, normalize
- Designing tables so that each fact is stored once: every entity gets its own table, and keys link the tables. Study card: Normalization →
- NULL
- A marker that a column has no value in this row. It is not zero and it is not an empty text. Study card: Relational data model →
O
- OLAP also: online analytical processing
- Online analytical processing: reading large volumes of historical data to report on it and to find patterns. Study card: Analytical workloads →
- OLTP also: online transactional processing
- Online transactional processing: recording many small business events, such as payments, quickly and reliably. Study card: Transactional workloads →
- OneLake
- The single data lake that all Microsoft Fabric workloads of an organization share, stored in Delta Lake format. Study card: Fabric and Azure Databricks →
P
- PaaS also: platform as a service
- Platform as a service: the provider runs the servers, the patching and the backups, and you use the service. Study card: Azure SQL family →
- page blob also: page blobs
- The blob type for random reads and writes. Azure uses it for virtual machine disks. Study card: Azure Blob storage →
- Parquet
- A columnar file format: values are stored by column, which makes analytical reads of a few columns fast. Study card: Data file formats →
- partition key
- The value that decides which partition an item is stored in. Items with the same partition key are kept together. Study card: Azure Table storage →
- pipeline also: pipelines
- A sequence of activities that moves and transforms data from its sources to its destinations. Study card: Data ingestion and processing →
- Power BI Desktop
- The free Windows application in which you connect to data, build the model and design reports. Study card: Power BI capabilities →
- Power BI service
- The online part of Power BI, where reports are published, refreshed on a schedule and shared. Study card: Power BI capabilities →
- primary key also: primary keys
- The column, or columns, whose value uniquely identifies each row of a table. Study card: Normalization →
Q
- Q&A visual
- A Power BI visual that answers a question typed in plain language with a chart. Study card: Data visualization types →
R
- relational database also: relational databases
- A database that stores data in tables and links those tables through key values. Study card: Common data stores →
- row key
- The value that identifies a row within its partition. Partition key and row key together are unique. Study card: Azure Table storage →
S
- SaaS also: software as a service
- Software as a service: a complete application that the provider runs and you only use. Study card: Fabric and Azure Databricks →
- scatter plot also: scatter plots
- A chart of points that shows how two numeric measures relate to each other. Study card: Data visualization types →
- schema also: schemas
- The agreed shape of the data: which fields exist and what type of value each one holds. Study card: Structured data →
- schema agnostic
- The database does not require items to share the same properties, so the shape of the data can vary and change. Study card: Cosmos DB use cases →
- schema on read
- No schema is enforced when data is written. The engine that queries the data defines the structure at that moment. Study card: Analytical data stores →
- semantic model also: semantic models
- The model behind Power BI reports: tables, relationships, measures and hierarchies, described in business terms. Study card: Power BI data models →
- semi-structured data also: semi-structured
- Data that has structure, but the fields can differ from one record to the next. JSON documents are the usual example. Study card: Semi-structured data →
- smart narrative also: smart narratives
- A Power BI visual that writes a text summary of the data and updates it when the data changes. Study card: Data visualization types →
- SMB also: Server Message Block
- Server Message Block: the file sharing protocol that works from Windows, Linux and macOS. Study card: Azure Files →
- snowflake schema
- A star schema in which a dimension table is split further into related detail tables. Study card: Analytical data stores →
- Spark Structured Streaming
- An Apache Spark library that treats a stream of data as a table that keeps growing. Study card: Real-time analytics services →
- SQL also: Structured Query Language
- Structured Query Language: the standard language for defining, querying and changing data in relational databases. Study card: SQL statement types →
- SQL Server on Azure VMs also: SQL Server on Azure Virtual Machines, SQL VM
- SQL Server installed on a virtual machine that you manage yourself: full control, and the most work. Study card: Azure SQL family →
- star schema also: star schemas
- A model with a fact table in the middle, linked to the dimension tables you group and filter by. Study card: Analytical data stores →
- stored procedure also: stored procedures
- SQL statements saved in the database under a name. An application runs them on command, often with parameters. Study card: Database objects →
- stream processing also: streaming
- Handling each event as it arrives, continuously, so the result is available within seconds or less. Study card: Batch versus streaming data →
- structured data also: structured
- Data with a fixed schema: every record has the same fields, usually shown as rows and columns. Study card: Structured data →
T
- throughput
- How much work a database can do per second. In Cosmos DB you set it on a database or a container. Study card: Cosmos DB use cases →
- Transact-SQL also: T-SQL
- Microsoft's dialect of SQL, used by SQL Server and the Azure SQL services. Study card: SQL statement types →
- transaction also: transactions
- A small unit of work that must succeed or fail as a whole, such as moving money between two accounts. Study card: Transactional workloads →
U
- Unity Catalog
- The governance layer of Azure Databricks: it controls access to data and tracks where data comes from. Study card: Fabric and Azure Databricks →
- unstructured data also: unstructured
- Data with no specific structure: documents, images, audio, video and other binary files. Study card: Unstructured data →
V
- virtual table
- A table that stores no data of its own: it shows the result of a query each time you use it. A view is one. Study card: Database objects →
W
- WHERE clause
- The part of a statement that limits it to the rows that match a condition. Study card: SQL statement types →