Azure Analytics Platforms

Tags intermediate guide azure cloud-platform

BenchBox supports Azure-native analytics platforms alongside multi-cloud options (Databricks, Snowflake).

Overview

Microsoft Azure offers multiple analytics platforms for different workload patterns:

Platform

Type

Focus Area

Status in BenchBox

Microsoft Fabric

Unified SaaS Platform

Data engineering, warehousing, BI, real-time analytics

Supported (Warehouse + Spark + Lakehouse SQL)

Azure Synapse Analytics

Enterprise Data Warehouse

Data warehousing, big data analytics

Supported (SQL + Spark)

Azure Data Explorer

Real-time Analytics Engine

Time-series, telemetry, streaming data

Under evaluation

Microsoft Fabric

Status: Supported (Warehouse + Spark + Lakehouse SQL)

Platform Type: Unified SaaS data analytics platform

Technical Characteristics:

  • OneLake data lake with multi-cloud support

  • Integrated workloads: Data Engineering, Data Factory, Data Science, Real-Time Intelligence, Data Warehouse, Databases

  • Capacity-based pricing model

  • Built-in Copilot AI assistance

  • Delta Lake Parquet format (not SQL Server native format)

Common use cases:

  • Organizations consolidating multiple data tools into unified platform

  • Teams requiring integrated data engineering, warehousing, and BI

  • Multi-cloud data lake architectures

  • Self-service BI with Power BI integration

BenchBox Integration:

  • Supports Microsoft Fabric Warehouse items via T-SQL

  • Entra ID authentication (service principal, default credential, interactive)

  • OneLake staging for bulk data loading via COPY INTO

  • Automatic query translation from standard SQL

BenchBox provides three adapters:

  • fabric-dw (or legacy fabric_dw) - For Warehouse items via T-SQL

  • fabric-spark - For Spark/Lakehouse workloads

  • fabric-lakehouse - For read-only SQL query benchmarking over Lakehouse tables

See Microsoft Fabric Platform Guide, Fabric Spark Platform Guide, and Fabric Lakehouse SQL Guide for detailed usage.

Azure Synapse Analytics

Status: Supported (SQL Pools + Spark)

Platform Type: Enterprise data warehouse and big data analytics service

Technical Characteristics:

  • Dedicated SQL Pools (MPP architecture)

  • Serverless SQL pools (pay-per-query)

  • Apache Spark pools for big data processing

  • T-SQL dialect with SQL Server compatibility

  • PolyBase for external data access

  • Azure AD authentication support

Common use cases:

  • Enterprise data warehousing on Azure

  • Big data analytics with Spark

  • Data lake queries via serverless SQL

  • Azure-native analytics workloads

Architecture: PaaS with choice of serverless or dedicated compute resources

BenchBox Integration:

  • synapse adapter for SQL pools via T-SQL

  • synapse-spark adapter for Spark pools

  • Entra ID authentication support

  • Automatic query translation from standard SQL

See Azure Synapse Spark Platform Guide for Spark-specific usage.

Azure Data Explorer (Kusto)

Platform Type: Fast analytics service for real-time data

Technical Characteristics:

  • Kusto Query Language (KQL) with T-SQL support

  • High-throughput ingestion (up to 12 Mbps per core)

  • Time-series analysis functions

  • Streaming data support

  • Integration with Azure Event Hubs, IoT Hub

Common use cases:

  • Real-time telemetry and log analytics

  • IoT data analysis

  • Time-series workloads

  • Application performance monitoring

Architecture: Fully managed service optimized for time-series and streaming data

Integration Considerations: KQL query translation, streaming ingestion patterns, time-series benchmark adaptations

Future Platforms

Azure Data Explorer is under evaluation for future BenchBox releases. Key considerations include:

  • KQL query translation requirements

  • Streaming ingestion patterns

  • Time-series benchmark adaptations