Skip to main content

Additional dashboards

You'll install the community system-tables demo so optional AI/BI dashboards run against your real billing data in ~15 min.

What you'll walk away with​

A set of AI/BI dashboards and notebooks running on your own billing data: consumption tracking, a DBU forecast, model-endpoint cost attribution, warehouse attribution, and UC/volume analysis. The intermediate tables land in a Unity Catalog catalog and schema you pick, which does not have to be main.

Official source (Demo Center): System Tables: Billing Forecast, Usage Analytics, and Access Auditing with Unity Catalog

tip

Confirm the demo slug is still system-tables on PyPI or GitHub before you standardize this in enablement.

Steps​

1. Install the dashboards from the dbdemos library​

Run these cells in a new Python notebook.

%pip install dbdemos
dbutils.library.restartPython()
import dbdemos

catalog = "mycatalog"
schema = "myschema"

dbdemos.install("uc-04-system-tables", catalog=catalog, schema=schema)

Set catalog and schema to UC assets you already have. Then wait until the notebook finishes and prints status for the notebooks and dashboards it created.

dbdemos install notebook showing the system tables demo setup with catalog and schema configuration

dbdemos install output listing the installed Databricks AI/BI dashboards and Genie Space for system tables

Verify​

Open the Dashboards and Genie Spaces sections and confirm the new assets are there.

Dashboards list showing the installed dbdemos System Tables dashboards in a Databricks workspace

Genie Spaces page showing the installed Databricks System Table Genie Space

Where people trip​

ModuleNotFoundError: dbdemos

Run %pip install dbdemos, then dbutils.library.restartPython(), before you import.

Permission denied writing catalog or schema

Pick a catalog and schema where your user can create tables, or ask a metastore admin to grant CREATE.

Permission denied on system tables

An admin has to grant SELECT on system.billing (and system.access for the audit-backed tiles).

Empty charts

New accounts have no history yet. Forecast tiles stay thin until enough daily usage accumulates.

Install hangs

Confirm the cluster has internet egress to PyPI and enough resources.

Next​