Create and use data dashboards

A dashboard generates and executes SQL over selected Catalog tables and displays chart results. Sources may span catalogs and databases. Access to a dashboard does not replace permissions on its underlying tables or SQL.

Create a dashboard and select its source

  1. Open dashboards, create a dashboard, and give it a recognizable name.

  2. In the data source selector, open the appropriate catalog and database, then select the tables to analyze.

  3. For cross-database analysis, select tables in other catalogs or databases, and check the final source list. Create a dashboard, name it, and select source tables by catalog and database. To analyze across databases, add tables from the other locations and review the complete source list before saving.

For example, combine orders from a sales database with products from another database. Check the join keys, units, and time scope first.

Add charts and check SQL

  1. Add a chart and describe the question.

  2. Review generated SQL, including qualified table names, joins, filters, and aggregation.

  3. Preview the SQL and inspect returned columns and sample values.

  4. Check chart settings and save.

  5. Run the chart and compare its results with the source data.

Generating a SQL draft does not execute it. Submitting a run does not prove results are ready.

Refresh and maintain

Refresh after source data changes, then inspect each chart’s loading, result, or error state. After changing sources, review dependent SQL and charts. Check inaccessible tables, changed columns, filters, and permissions before retrying.

If the first chart keeps loading, check whether it already exists before creating another. For empty charts, inspect the SQL result before changing presentation settings.

View results, drill through, and export CSV

Each chart card supports inspecting and exporting underlying data:

  • View query result & export CSV: Open View query result from a chart’s actions to inspect returned rows and click Export CSV.

  • Detail drill-through & SQL handoff: Click a dimension or data point to open the Detail rows drawer with the generated Detail SQL. Drill-through queries run with your current Effective Role permissions; click Export current details to download visible rows or Open in SQL Editor to continue exploring in SQL Editor.

View historical snapshots

Use the View time control at the top of a dashboard (default: Now):

  1. Select a preset (1 day, 7 days, 15 days, 30 days, 3 months, 6 months, or 1 year) or choose Custom to pick a specific timestamp.

  2. If a snapshot exists at that time, charts display the historical snapshot; if the UI reports no snapshot is available, select another timestamp or switch back to Now.

  3. Historical snapshots do not support detail drill-through; switch back to Now before drilling down.

Configure chart alerts and view records

Metric cards, Line charts, Bar charts, and Area charts support dashboard chart alerts (other chart types show alert configuration as unsupported):

  1. When adding or choosing Edit on a supported chart, turn on Enable data alert.

  2. Set Evaluation data, Operator, and Threshold.

  3. Choose Alert level (Critical, Warning, Info) and Notification interval (Realtime, Every 15 minutes, Every hour, Every 6 hours, or Every 24 hours), then save.

  4. Click Alert records in the dashboard header to review triggered chart, condition, data point, level, status (Alerting / Recovered), and timestamp.

Common problems

Symptom

How to check

The first chart keeps loading

Check current run state and errors, then refresh to see whether a chart already exists to avoid duplicates.

SQL cannot be previewed

Check qualified database/table names, selected sources, fields, and access privileges. Retain errors shown by the page.

Charts contain no data

Check SQL results with the same conditions first, then inspect chart columns and time filters.

Refresh fails

Check that sources remain available and SQL still matches current table structures.

API, SDK, and CLI

Last updated on