v5

latestOpenAPI 3.1.02026-08-081,0211,3082.1 MB

DATABRICKS

Databricks

Databricks Warehouses

Get time-bucketed query metrics timeline for a warehouse

Get the time-bucketed metrics timeline powering the SQL warehouse Monitoring page's stacked charts (latency, queueing, scan size, cache hit rate, spill).

Hybrid-sourced: historical buckets come from the ClickHouse mirror, and the fresh tail the mirror hasn't ingested yet (up to ~24h) is read live from Databricks system tables via the Statement Execution API against the tenant's ingestion SQL warehouse — this will auto-start that warehouse if it is currently suspended. A live-tail failure degrades to a ClickHouse-only response rather than erroring.

The lookback window is capped at MAX_MONITORING_LOOKBACK_DAYS days regardless of what's requested — end_date is clamped to today and start_date is clamped to end_date - MAX_MONITORING_LOOKBACK_DAYS (silently, not a 400; the frontend's date-range picker clamps to the same bound).

get/api/v1/databricks/warehouses/{warehouse_id}/query-metrics-timeline

Path parameters

warehouse_idstring required

Query parameters

start_datestring date required

Start date

end_datestring date

End date

granularitystring

Time granularity: 'hour' or 'day'

instance_idinteger nullable

Databricks instance id

navigationSourcestring nullable

Headers

x-tenantstring required

Response

Successful Response

warehouse_idstring required

granularitystring required

ch_data_throughstring nullable

live_tail_usedboolean