Available Advisors¶

The Fabric Warehouse Advisor ships with modular advisor modules. Each one analyses a different aspect of warehouse health and produces its own reports.

Advisor Comparison¶

	Data Clustering	Performance Check
Purpose	Recommend optimal `CLUSTER BY` columns	Detect performance anti-patterns
Output model	Scored recommendations (0–100)	Findings (Critical / High / Medium / Low / Info)
Applies to	DataWarehouse only	DataWarehouse and Lakehouse SQL Endpoints
Config class	`DataClusteringConfig`	`PerformanceCheckConfig`
Result class	`DataClusteringResult`	`PerformanceCheckResult`
Reports	Text, Markdown, HTML	Text, Markdown, HTML
DDL generation	Yes (CTAS statements)	Yes (per-finding SQL fixes)
Phases	7 (metadata → scoring)	7 (edition → statistics)

Data Clustering Advisor¶

Analyses your actual query patterns (via Query Insights), combines them with table metadata and column cardinality estimates, and scores every candidate column from 0 to 100. You get a clear report telling you exactly what to cluster and why.

from fabric_warehouse_advisor import DataClusteringAdvisor, DataClusteringConfig

config = DataClusteringConfig(warehouse_name="MyWarehouse")
result = DataClusteringAdvisor(spark, config).run()
displayHTML(result.html_report)

Full documentation →

Performance Check Advisor¶

Scans for common performance pitfalls across multiple categories, including warehouse edition detection, data‑type anti‑patterns, caching configuration, collation settings, query regression, statistics health, and V‑Order optimization state.

Checking Warehouse edition

Not all checks apply to every warehouse edition.

from fabric_warehouse_advisor import PerformanceCheckAdvisor, PerformanceCheckConfig

config = PerformanceCheckConfig(warehouse_name="MyWarehouse")
result = PerformanceCheckAdvisor(spark, config).run()
displayHTML(result.html_report)

Full documentation →

Running Both Advisors Together¶

For a comprehensive warehouse health assessment, run both advisors in the same notebook:

from fabric_warehouse_advisor import (
    DataClusteringAdvisor, DataClusteringConfig,
    PerformanceCheckAdvisor, PerformanceCheckConfig,
)

warehouse = "MyWarehouse"

# Performance check first — fix anti-patterns before optimizing clustering
pc_result = PerformanceCheckAdvisor(spark, PerformanceCheckConfig(
    warehouse_name=warehouse,
)).run()

dc_result = DataClusteringAdvisor(spark, DataClusteringConfig(
    warehouse_name=warehouse,
)).run()

# Display both reports
displayHTML(pc_result.html_report)
displayHTML(dc_result.html_report)

Recommended order

Run the Performance Check first — resolving data-type issues, enabling caching, and fixing statistics will improve the accuracy of the Data Clustering advisor's recommendations.