Health, Diagnostics & Performance

The procedures in this section help validate cube health, collect diagnostic information, review licensing capacity, analyze query performance, and evaluate physical optimization recommendations.

HEALTH_CHECK

Minimum role: CUBE_PUBLIC

HEALTH_CHECK(P_CUBE_NAME) -> TABLE

Validates cube configuration, structures, and formulas.

Parameters

Parameter	Type	Req/Opt	Description
`P_CUBE_NAME`	`VARCHAR`	opt	Cube to check. Pass `NULL` to check every cube.

Returns

Returns:

CUBE_NAME
DIM_NAME
CHECK_TYPE
STATUS
MESSAGE

Checks include DIMENSION_TABLE, DIMENSION_ROW_COUNT, INPUT_TABLE_CONFIG, QUERY_FUNCTION, SHARED_TABLE, DATA_FRESHNESS, CONFIG_INTEGRITY, FORMULA_SYNTAX, FORMULA_BRACKETS, FORMULA_CASE_END, FORMULA_EMPTY, FORMULA_NO_DIMENSION, FORMULA_DIVISION_SAFETY, FORMULA_ESSBASE_SYNTAX, FORMULA_COUNT, and STALE_AI_AGENT.

Example

Run a health check:

CALL CUBE.HEALTH_CHECK('FINANCE');

Show only non-OK results:

SELECT *
FROM TABLE(RESULT_SCAN(LAST_QUERY_ID()))
WHERE STATUS <> 'OK';

GENERATE_SUPPORT_INFO

Minimum role: CUBE_ADMIN

GENERATE_SUPPORT_INFO(P_CUBE_NAME) -> VARIANT

Collects version, configuration, and application state into a diagnostic bundle.

Run this procedure when gathering information for a support request.

Parameters

Parameter	Type	Req/Opt	Description
`P_CUBE_NAME`	`VARCHAR`	opt	Cube name. Pass `NULL` for system-level information.

Example

CALL CUBE.GENERATE_SUPPORT_INFO('FINANCE');

GET_PLAN_INFO

Minimum role: CUBE_ADMIN

GET_PLAN_INFO() -> OBJECT

Reports the licensing position of the current installation.

This procedure takes no parameters.

Returns

Returns:

ACCOUNT_IDENTIFIER
ACCOUNT_NAME
ORGANIZATION_NAME
PLAN_TIER
CUBES_ALLOWED
CUBES_USED
CUBES_REMAINING
SUPPORT_CONTACT

Check CUBES_REMAINING before a migration that will create multiple cubes. Cube creation fails after the installation’s cube allowance has been used.

Example

CALL CUBE.GET_PLAN_INFO();

PERFORMANCE_REPORT

Minimum role: CUBE_ADMIN

PERFORMANCE_REPORT(
    P_CUBE_NAME,
    P_DAYS
) -> VARIANT

Analyzes query performance for a cube over a specified period.

Parameters

Parameter	Type	Req/Opt	Description
`P_CUBE_NAME`	`VARCHAR`	req	Cube name.
`P_DAYS`	`VARCHAR`	opt	Number of days to analyze.

Example

CALL CUBE.PERFORMANCE_REPORT(
    'FINANCE',
    '30'
);

SLOW_QUERY_INSPECTOR

Minimum role: CUBE_ADMIN

SLOW_QUERY_INSPECTOR(
    P_CUBE_NAME,
    P_LIMIT,
    P_DAYS,
    P_SLOW_MS
) -> VARIANT

Identifies the slowest queries and suggests changes that may improve performance.

Parameters

Parameter	Type	Req/Opt	Description
`P_CUBE_NAME`	`VARCHAR`	opt	Cube name. Pass `NULL` for all cubes.
`P_LIMIT`	`VARCHAR`	opt	Maximum number of queries to return.
`P_DAYS`	`VARCHAR`	opt	Lookback period in days.
`P_SLOW_MS`	`VARCHAR`	opt	Threshold in milliseconds for a query to be considered slow.

Example

CALL CUBE.SLOW_QUERY_INSPECTOR(
    'FINANCE',
    '20',
    '7',
    '5000'
);

RECOMMEND_CLUSTERING

Minimum role: CUBE_ADMIN

RECOMMEND_CLUSTERING(
    P_CUBE_NAME,
    P_DATA_TABLE,
    P_MIN_QUERIES,
    P_APPLY
) -> VARIANT

Analyzes actual query history and recommends clustering keys for cube data tables.

Parameters

Parameter	Type	Req/Opt	Description
`P_CUBE_NAME`	`VARCHAR`	opt	Cube name.
`P_DATA_TABLE`	`VARCHAR`	opt	Specific data table to analyze.
`P_MIN_QUERIES`	`VARCHAR`	opt	Minimum query count before making a recommendation.
`P_APPLY`	`VARCHAR`	opt	Pass `'TRUE'` to apply the recommendations. Omit to report only.

By default, the procedure reports recommendations without changing anything.

Example

CALL CUBE.RECOMMEND_CLUSTERING(
    'FINANCE',
    NULL,
    NULL,
    NULL
);

RECOMMEND_SEARCH_OPTIMIZATION

Minimum role: CUBE_ADMIN

RECOMMEND_SEARCH_OPTIMIZATION(
    P_CUBE_NAME,
    P_DATA_TABLE,
    P_MIN_QUERIES,
    P_APPLY
) -> VARIANT

Analyzes actual query history and recommends Snowflake search optimization for cube data tables.

Parameters

Parameter	Type	Req/Opt	Description
`P_CUBE_NAME`	`VARCHAR`	opt	Cube name.
`P_DATA_TABLE`	`VARCHAR`	opt	Specific data table to analyze.
`P_MIN_QUERIES`	`VARCHAR`	opt	Minimum query count before making a recommendation.
`P_APPLY`	`VARCHAR`	opt	Pass `'TRUE'` to apply the recommendations. Omit to report only.

By default, the procedure reports recommendations without changing anything.

Example

CALL CUBE.RECOMMEND_SEARCH_OPTIMIZATION(
    'FINANCE',
    NULL,
    NULL,
    NULL
);