Infrastructure

Execution History

A complete audit log of every query executed through Datavine — who ran it, when, how long it took, how much data was scanned, and whether it was served from cache.

What gets logged

Every query execution — whether from the SQL Editor, Data Chat, Dashboard Builder, or AI-generated — is recorded with full context:

SQL ContentThe complete query text that was executed. Syntax-highlighted in the detail view.

StatusSuccess, failed, or cancelled. Failed queries include the error message from BigQuery.

DurationWall-clock execution time in milliseconds. Cached queries show near-zero duration.

Rows ReturnedNumber of result rows. Useful for spotting unexpectedly large or empty result sets.

Bytes ScannedTotal data processed by BigQuery. Directly correlates with cost ($6.25/TB on-demand).

Cache HitWhether results came from Datavine’s query cache (free) or were executed against BigQuery (billed).

Executed ByWhich user ran the query, or ‘system’ for scheduled/automated executions.

SourceWhere the query originated: SQL Editor, Data Chat, Dashboard widget, Anomaly Scan, or API.

TimestampWhen the query was submitted and when results were returned.

Filtering and search

The execution log supports filtering by status (success/failed/cached), date range, user, and source. Search by SQL content to find specific queries. Sort by duration, bytes scanned, or timestamp to identify expensive or slow queries.

Re-running queries

Click any execution row to see full details. From there, click Run Again to load the SQL back into the SQL Editor and re-execute it, or Copy SQL to grab the query text.

Cost attribution

Each execution shows its estimated cost based on bytes scanned. Aggregate views let you see total cost per user, per day, or per source — helping identify who or what is driving BigQuery spend.

Query Caching

Issue Boards