6. Comprehensive Glossary

A

Accelerated Shortcut: Shortcut mode in Real-Time Intelligence that caches data in Eventhouse for fast KQL queries. (See: 3.2.3)

ACID: Atomicity, Consistency, Isolation, Durability—transaction properties ensuring data integrity. (See: 2.1.1)

Autoscale: Automatic adjustment of node count based on workload demand. (See: 2.2.1)

C

Capacity: The compute and storage engine behind Fabric workloads. (See: 2.1.2)

CTAS: Create Table As Select—T-SQL statement creating a new table from query results. (See: 3.3.3)

Column-Level Security (CLS): Restricts access to specific columns for unauthorized users. (See: 2.4.2)

D

Data Access Role: OneLake security feature granting folder-level and file-level access within a lakehouse. (See: 2.4.6)

Data Integration Unit (DIU): Compute measure for Copy activity throughput in pipelines. (See: 4.3.1)

Data Workflow Settings: Workspace-level configuration controlling Dataflow Gen2 concurrency and compute allocation. (See: 2.2.2)

Dataflow Gen2: Low-code ETL tool using Power Query (M) engine. (See: 2.5.1)

Delta Lake: Open-source storage layer providing ACID transactions on data lakes. (See: 2.1.1)

Deployment Pipeline: Automated content promotion through development stages. (See: 2.3.3)

Deployment Rules: Configuration overrides per deployment stage. (See: 2.3.3)

Domain: Logical grouping of workspaces for enterprise governance. (See: 2.1.2)

Dynamic Allocation: Spark feature adjusting executor count based on workload. (See: 2.2.1)

Dynamic Data Masking (DDM): Obscures sensitive data in query results without modifying stored data. (See: 2.4.3)

E

Eventstream: Fabric item for real-time event ingestion and routing. (See: 3.4.2)

Event Processor: No-code transformation component within eventstreams. (See: 3.4.2)

Eventhouse: Fabric item containing KQL databases for real-time analytics. (See: 3.4.1)

F

Fail Activity: Pipeline activity that terminates execution with custom error message. (See: 4.2.1)

Full Load: Loading pattern that extracts complete dataset every run. (See: 3.1.1)

G

Gateway: On-premises data gateway for connecting to local data sources. (See: 4.2.5)

Git Integration: Version control for Fabric items using Azure DevOps or GitHub. (See: 2.3.1)

I

Incremental Load: Loading pattern that extracts only new or modified records. (See: 3.1.1)

K

KQL: Kusto Query Language—query language for real-time analytics. (See: 3.4.4)

KQL Database: Fabric item optimized for time-series and streaming data. (See: 3.4.1)

L

Lakehouse: Fabric item combining data lake and warehouse capabilities. (See: 2.1.3)

Log Analytics Integration: Feature to export Fabric workspace logs to Azure Log Analytics for advanced querying and compliance. (See: 4.1.4)

M

Managed Private Endpoint: Private network connection from Fabric to Azure services using Azure Private Link. (See: 2.4.5)

Memory-Optimized Nodes: Spark nodes with higher memory ratio for data-intensive operations. (See: 2.2.1)

Mirroring: Replication of external data into Fabric (database or metadata). (See: 3.2.4)

Monitor Hub: Centralized monitoring interface for all Fabric activities. (See: 4.1.1)

N

Native Execution Engine: Fabric's optimized Spark engine (doesn't support UDFs). (See: 4.3.3)

O

OneLake: Unified data lake for all Fabric workloads. (See: 2.1.1)

OPTIMIZE: Delta Lake command to consolidate small files. (See: 4.3.2)

P

Pipeline: Orchestration item for coordinating data movement and transformation. (See: 2.5.1)

Power Query (M): Functional language for data transformation in Dataflows. (See: 3.3.1)

R

Real-Time Hub: Monitoring interface for streaming data flows. (See: 4.1.3)

Result Set Caching: Automatic caching of warehouse query results for faster repeated queries. (See: 4.3.4)

Retention Policy: Eventhouse policy controlling automatic data deletion after a time period. (See: 4.3.5)

Row-Level Security (RLS): Filters rows based on user identity. (See: 2.4.2)

S

SCD (Slowly Changing Dimension): Patterns for handling dimension changes over time. (See: 3.1.3)

Sensitivity Labels: Classification metadata from Microsoft Purview. (See: 2.4.4)

Session Tags: Identifiers enabling Spark session reuse across activities. (See: 2.2.1)

Shortcut: Virtual pointer to external data without copying. (See: 3.2.3)

Spark Structured Streaming: Spark API for processing streaming data. (See: 3.4.3)

Starter Pool: Pre-warmed Spark cluster for fast session start. (See: 2.2.1)

Subdomain: Child grouping within a domain for finer governance. (See: 2.1.2)

T

TRY/CATCH: T-SQL error handling construct. (See: 4.2.3)

Trusted Workspace Access: Configuration allowing Fabric workspaces to bypass Azure Storage firewall rules. (See: 2.4.5)

Tumbling Window: Fixed, non-overlapping time intervals for aggregation. (See: 3.4.5)

V

V-Order: Microsoft's columnar optimization for Parquet files. (See: 4.3.2)

W

Watermark: Tracking value for incremental load progress. (See: 3.1.1)

Windowing Functions: Functions that divide streams into finite chunks for aggregation. (See: 3.4.5)

Workspace: Container for Fabric items with defined access roles. (See: 2.1.2)

Workspace Logging: Export of Fabric activity logs to Azure Log Analytics for compliance and advanced monitoring. (See: 4.1.4)