fabric-warehouse-advisor 1.1.3

A modular advisory framework for Microsoft Fabric Warehouse — Data Clustering, Performance, Security and more.

Fabric Warehouse Advisor

A modular Python advisory framework for Microsoft Fabric Warehouse. Each advisor module analyses a different aspect of warehouse health and produces scored recommendations with rich reports.

Available Advisors:

• Data Clustering Advisor — assesses which tables and columns should use Data Clustering (scored 0–100).
• Performance Check Advisor — scans for data-type anti-patterns, caching misconfigurations, stale statistics, and V-Order issues (findings-based).
• Security Check Advisor — analyses permissions, roles, RLS, CLS, and Dynamic Data Masking configuration (findings-based).

It runs entirely inside a Fabric Notebook. The Microsoft Fabric Data Warehouse connector comes pre-installed in the Fabric runtime, and Query Insights is enabled by default on every warehouse. A Lakehouse is required only when the solution is installed from a wheel file stored in OneLake.

Installation

To install Fabric Warehouse Advisor, run:

%pip install fabric-warehouse-advisor

For version information, dependencies, and release notes, see the details.

Quick Start

Data Clustering

from fabric_warehouse_advisor import DataClusteringAdvisor, DataClusteringConfig

config = DataClusteringConfig(
    warehouse_name="MyWarehouse",
)

advisor = DataClusteringAdvisor(spark, config)
result = advisor.run()

# Rich HTML report — best way to view results in a Fabric notebook
displayHTML(result.html_report)

Performance Check

from fabric_warehouse_advisor import PerformanceCheckAdvisor, PerformanceCheckConfig

config = PerformanceCheckConfig(
    warehouse_name="MyWarehouse",
)

advisor = PerformanceCheckAdvisor(spark, config)
result = advisor.run()

displayHTML(result.html_report)

Security Check

from fabric_warehouse_advisor import SecurityCheckAdvisor, SecurityCheckConfig

config = SecurityCheckConfig(
    warehouse_name="MyWarehouse",
)

advisor = SecurityCheckAdvisor(spark, config)
result = advisor.run()

displayHTML(result.html_report)

Screenshots

Each advisor produces a rich, interactive HTML report with light and dark themes.

Data Clustering

Security Check

Performance Check

Documentation

Document	Description
Getting Started	Installation, first run, working with results
Advisors Overview	Comparison of all available advisors
Data Clustering
Overview	Analyzes query patterns, table metadata, and column cardinality to identify and score the best candidate columns for data clustering, optimizing physical data organization on OneLake for better query speed.
Performance Check
Overview	Identifies common performance pitfalls in Fabric Warehouses and Lakehouse SQL Endpoints by auditing data types, caching status, V-Order optimization, statistics health, and query performance regressions.
Security Check
Overview	Scans Microsoft Fabric Warehouses for security misconfigurations, covering schema permissions, custom roles, Row-Level Security (RLS), Column-Level Security (CLS), and Dynamic Data Masking to provide actionable findings and SQL fixes.

Acknowledgements

Report icons provided by Flaticon:

• Cyber security icons created by Freepik - Flaticon
• Performance icons created by Freepik - Flaticon
• Graph icons created by Karacis - Flaticon

License

MIT — see LICENSE for details.