Metabase Overview¶

How Dango auto-configures Metabase for your DuckDB warehouse.

Overview¶

Metabase is the built-in BI tool in Dango, providing dashboards, SQL queries, and visualizations for your data.

What Dango automates:

• Metabase installation via Docker
• DuckDB connection setup
• Admin user creation
• Database schema sync

What Metabase provides:

• Visual query builder
• SQL editor
• Interactive dashboards
• Chart visualizations

Quick Start¶

Access Metabase¶

Start your Dango platform:

dango start

Metabase is available at: http://localhost:3000

Auto-Created Credentials¶

Dango automatically creates an admin user:

• Email: admin@dango.local
• Password: dangolocal123

First Login¶

1. Open http://localhost:3000
2. Login with credentials above
3. Start querying - DuckDB connection is pre-configured

Auto-Configuration¶

DuckDB Connection¶

Dango pre-configures the database connection:

• Database Name: DuckDB
• Database File: data/warehouse.duckdb (shared volume)
• Connection: Automatic (no setup required)

Available Schemas¶

All Dango data layers are accessible in Metabase:

Browse in Metabase: Home → Browse data → DuckDB

Database Sync¶

Metabase syncs schema metadata:

• On startup: Full schema scan
• Daily: Automatic re-sync (2 AM default)
• Manual: Admin → Databases → DuckDB → "Sync database schema now"

Querying Dango Data¶

Browse Tables¶

Navigate to: Home → Browse data → DuckDB

DuckDB
├── raw_stripe
│   ├── charges
│   ├── customers
│   └── subscriptions
├── staging
│   └── stg_stripe_charges
└── marts
    └── customer_metrics

Click any table to preview data and create questions.

SQL Queries¶

Open the SQL editor: Home → SQL

Example - Query marts table:

SELECT
    DATE_TRUNC('month', created_at) as month,
    COUNT(*) as customer_count,
    SUM(lifetime_value) as total_revenue
FROM marts.customer_metrics
GROUP BY month
ORDER BY month DESC

Example - Query with dlt metadata:

SELECT *
FROM raw_stripe.charges
WHERE _dlt_extracted_at > CURRENT_DATE - INTERVAL 1 DAY
ORDER BY _dlt_extracted_at DESC

Best Practice: Query Marts¶

For dashboards, query pre-aggregated marts instead of raw tables:

-- Good: Query marts (fast)
SELECT * FROM marts.revenue_by_month

-- Avoid: Aggregate raw data in Metabase (slow)
SELECT DATE_TRUNC('month', created), SUM(amount)
FROM raw_stripe.charges
GROUP BY 1

Create aggregations in dbt, not Metabase.

Data Refresh Workflow¶

Update Data in Metabase¶

Metabase queries live DuckDB data. To refresh:

# 1. Sync new source data
dango sync

# 2. Run dbt transformations
dango run

# 3. Reload Metabase dashboard (no action needed!)

Metabase shows updated data immediately after dango run completes.

Schema Changes¶

When you add new dbt models:

# 1. Create model
dango model add

# 2. Run transformations
dango run

# 3. Sync Metabase schema
# Admin → Databases → DuckDB → "Sync database schema now"

Or wait for the daily automatic sync.

Metabase vs dbt Docs¶

Use both:

• dbt Docs: Understand data lineage and transformations
• Metabase: Build dashboards and share insights

Troubleshooting¶

Cannot Connect to Metabase¶

Check if running:

docker ps | grep metabase

Restart:

dango stop
dango start

Schema Not Updating¶

Trigger manual sync:

1. Admin → Databases → DuckDB
2. Click "Sync database schema now"

Missing Tables¶

Verify table exists:

duckdb data/warehouse.duckdb "SHOW TABLES FROM marts;"

If the table exists but doesn't appear in Metabase, trigger a manual sync.

Query Timeout¶

For complex queries:

1. Create a dbt mart with pre-aggregated data
2. Query the mart instead of raw tables
3. Optionally increase timeout: Admin → Databases → DuckDB → Advanced options

Next Steps¶