Configuration Reference¶

All parameters are fields of the PerformanceCheckConfig dataclass. Create an instance, override the defaults you need, and pass it to PerformanceCheckAdvisor.

from fabric_warehouse_advisor import PerformanceCheckAdvisor, PerformanceCheckConfig

config = PerformanceCheckConfig(
    warehouse_name="MyWarehouse",
    check_vorder=True,
    stale_stats_threshold_days=14,
    verbose=True,
)

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

Connection Parameters¶

Parameter	Type	Default	Description
`warehouse_name`	`str`	`""`	Required. The Fabric Warehouse or Lakehouse SQL Endpoint name.
`workspace_id`	`str`	`""`	Workspace GUID. Only needed for cross-workspace access.
`warehouse_id`	`str`	`""`	Warehouse item GUID. Only needed for cross-workspace access.
`sql_endpoint_id`	`str`	`""`	SQL Endpoint item GUID. For cross-workspace access to a Lakehouse SQL Analytics Endpoint.

Scope Filtering¶

Parameter	Type	Default	Description
`schema_names`	`list[str]`	`[]`	Restrict analysis to specific schemas. Empty = all user schemas.
`table_names`	`list[str]`	`[]`	Restrict analysis to specific tables. Each entry can be `"table_name"` (any schema) or `"schema.table_name"`. Empty = all tables.

Examples:

# Only check tables in the 'sales' schema
config = PerformanceCheckConfig(
    warehouse_name="MyWarehouse",
    schema_names=["sales"],
)

# Only check specific tables
config = PerformanceCheckConfig(
    warehouse_name="MyWarehouse",
    table_names=["dbo.FactSales", "dbo.DimCustomer"],
)

Check Category Toggles¶

Each check category can be independently enabled or disabled:

Parameter	Type	Default	Description
`check_data_types`	`bool`	`True`	Enable data-type anti-pattern analysis.
`check_caching`	`bool`	`True`	Enable caching configuration analysis.
`check_statistics`	`bool`	`True`	Enable statistics health analysis.
`check_vorder`	`bool`	`True`	Enable V-Order state check. Auto-skipped for LakeWarehouse.
`check_collation`	`bool`	`True`	Enable collation mismatch check.
`check_query_regression`	`bool`	`True`	Enable query regression detection.

Example — run only the statistics check:

config = PerformanceCheckConfig(
    warehouse_name="MyWarehouse",
    check_data_types=False,
    check_caching=False,
    check_vorder=False,
    check_statistics=True,
    check_collation=False,
    check_query_regression=False,
)

Data Type Thresholds¶

Fine-grained control over which data-type patterns are flagged:

Parameter	Type	Default	Description
`varchar_max_warning`	`bool`	`True`	Flag `VARCHAR(MAX)` columns.
`char_to_varchar_warning`	`bool`	`True`	Flag `CHAR(n)` columns where `VARCHAR(n)` may be better.
`nullable_warning`	`bool`	`True`	Flag nullable columns that could be `NOT NULL`.
`oversized_varchar_threshold`	`int`	`8000`	Flag `VARCHAR(n)` at or above this length.
`decimal_over_precision_threshold`	`int`	`18`	Flag `DECIMAL` / `NUMERIC` with precision above this value.
`datetime_as_string_warning`	`bool`	`True`	Flag string columns whose names suggest date/time data.
`float_for_money_warning`	`bool`	`True`	Flag `FLOAT` / `REAL` on monetary-sounding columns.
`bigint_where_int_suffices_warning`	`bool`	`True`	Flag `BIGINT` on columns suggesting small-range values.

Example — only flag critical issues:

config = PerformanceCheckConfig(
    warehouse_name="MyWarehouse",
    varchar_max_warning=True,       # CRITICAL: keep on
    char_to_varchar_warning=False,  # suppress CHAR warnings
    nullable_warning=False,         # suppress nullable INFO
)

Caching Settings¶

Parameter	Type	Default	Description
`result_cache_check`	`bool`	`True`	Check whether result set caching is enabled.
`cold_start_analysis`	`bool`	`True`	Analyse cold-start patterns from Query Insights.
`cold_start_lookback_hours`	`int`	`24`	How many hours back to look in `exec_requests_history`.
`cache_hit_ratio_warning_threshold`	`float`	`0.3`	Warn if the cache hit ratio is below this fraction (30%).

Statistics Settings¶

Parameter	Type	Default	Description
`stale_stats_threshold_days`	`int`	`7`	Warn if statistics are older than this many days.
`row_drift_pct_threshold`	`float`	`20.0`	Warn if actual row count differs from stats estimate by more than this %.
`proactive_refresh_check`	`bool`	`True`	Check if proactive statistics refresh is enabled.
`tables_without_stats_check`	`bool`	`True`	Check for tables that have no statistics at all.

Query Regression Settings¶

Parameter	Type	Default	Description
`regression_lookback_days`	`int`	`7`	Window (days) defining the "recent" period. Baseline is the preceding historical range within Query Insights' 30-day retention.
`regression_factor_warning`	`float`	`2.0`	Flag a query as WARNING when its recent median duration is at least this many times the baseline median.
`regression_factor_critical`	`float`	`5.0`	Flag a query as CRITICAL when its recent median duration is at least this many times the baseline median.
`regression_min_executions`	`int`	`3`	Minimum executions required in both baseline and recent windows before a query is compared.

V-Order Settings¶

Parameter	Type	Default	Description
`vorder_warn_if_disabled`	`bool`	`True`	Flag if V-Order is disabled on the warehouse.

Output¶

Parameter	Type	Default	Description
`verbose`	`bool`	`False`	Print intermediate debug output for each phase.

Validation¶

The config is validated automatically when advisor.run() is called:

warehouse_name must be set to a non-empty value (not the placeholder "<your_warehouse_name>")

If the check fails, a ValueError is raised with a descriptive message.