Skip to content

Dashboard Provisioning

Provision a pre-built monitoring dashboard in Metabase with a single command. The "Data Pipeline Health" dashboard gives you instant visibility into sync status, data freshness, test results, and row count trends.

Overview

dango dashboard provision creates a Metabase dashboard with 6 visualizations covering pipeline health. It authenticates with Metabase, creates individual cards (questions), assembles them into a dashboard with a grid layout, and returns the dashboard URL.

Quick Start

dango dashboard provision

Uses http://localhost:3000 by default.

dango dashboard provision --url https://your-domain.com

Provide the public URL of your Metabase instance.

You'll be prompted for the admin password:

Password: ********

✓ Dashboard provisioned successfully!

  Dashboard: Data Pipeline Health
  URL: http://localhost:3000/dashboard/1
  Cards created:
    • Pipeline Health Score
    • dbt Test Results
    • Data Sources Overview
    • Sync History (Last 7 Days)
    • Row Counts Over Time
    • Data Freshness by Source

Command Options

dango dashboard provision [--url URL] [--username EMAIL] [--password PASSWORD]
Option Default Description
--url http://localhost:3000 Metabase instance URL
--username Auto-detected Metabase admin email
--password Prompted Metabase admin password (entered interactively)

Username Resolution

The admin username is resolved in this order:

  1. Explicit --username flag
  2. DANGO_ADMIN_EMAIL environment variable
  3. Active admin user from the auth database
  4. Fallback: [email protected]

Dashboard Contents

The provisioned dashboard contains 6 cards arranged in a grid:

Layout

┌──────────────────┬──────────────────┬──────────────────┐
│ Pipeline Health  │ dbt Test Results │ Data Sources     │
│ Score (gauge)    │ (scalar)         │ Overview (table) │
│ 6×4              │ 6×4              │ 6×4              │
├────────────────────────────┬────────────────────────────┤
│ Sync History 7 Days       │ Row Counts Over Time       │
│ (line chart) 9×6          │ (area chart) 9×6           │
├────────────────────────────┴────────────────────────────┤
│ Data Freshness by Source (table) 18×4 — full width     │
└─────────────────────────────────────────────────────────┘

Cards

Card Visualization Position Description
Pipeline Health Score Gauge Row 0, Col 0 (6×4) Overall health score 0-100
dbt Test Results Scalar Row 0, Col 6 (6×4) Count of failed tests vs total
Data Sources Overview Table Row 0, Col 12 (6×4) Source name, type, enabled status
Sync History (Last 7 Days) Line chart Row 4, Col 0 (9×6) Sync activity per day
Row Counts Over Time Area chart Row 4, Col 9 (9×6) Data growth trend over 30 days
Data Freshness by Source Table Row 10, Col 0 (18×4) Row count and last update per source

Each card executes a native SQL query against the DuckDB database connected to Metabase.

Customizing the Dashboard

After provisioning, the dashboard is a standard Metabase dashboard. You can:

  • Edit the layout — drag and resize cards in Metabase
  • Add cards — create new questions and add them to the dashboard
  • Modify queries — click any card to edit its underlying SQL
  • Add filters — add interactive filter widgets to the dashboard
  • Change visualizations — switch chart types (e.g., line to bar)

Save Your Customizations

After customizing the dashboard in Metabase, run dango metabase save to export it as YAML. This lets you version-control your changes and restore them later. See Save & Load.

Re-Provisioning

Running dango dashboard provision again creates a new dashboard — it does not update the existing one. If you want a fresh dashboard:

  1. Delete the old dashboard in Metabase
  2. Run dango dashboard provision again

For updating an existing dashboard, edit it directly in the Metabase UI instead.

Troubleshooting

Authentication Failed

Error: Authentication failed. Check your credentials.

Verify your Metabase admin email and password. The Metabase admin password was auto-generated during dango start and stored in .dango/metabase.yml. This is separate from the Dango admin password set during dango init.

# Check which admin email is configured
dango auth list-users

dango auth list-users
# Or check the environment variable
echo $DANGO_ADMIN_EMAIL

DuckDB Database Not Found

Error: No DuckDB database configured in Metabase.

Metabase needs a DuckDB connection configured during setup. This is normally done automatically by dango start. If the connection is missing:

  1. Verify Metabase is running: dango status
  2. Check if DuckDB is connected in the Metabase admin panel under Admin > Databases
  3. Re-run setup if needed: dango start

Empty Dashboard

The dashboard cards show "No results" if there's no data in DuckDB yet:

  1. Run dango sync to load data
  2. Run dango run to build dbt models
  3. Run dango metabase refresh to sync the schema
  4. Open the dashboard — cards should now show data

Next Steps