Quick Start Guide

This guide walks you through getting LakeSentry running — from creating your account to viewing your first cost breakdown. The whole process takes about 15 minutes of active setup, plus a short wait for initial data collection.

Prerequisites

Before you start, make sure you have:

• A Databricks account with Unity Catalog enabled
• A service principal with read access to system tables, or a personal access token for quick evaluation
• At least one SQL warehouse available in your Databricks workspace

Step 1: Create your LakeSentry account

1. Go to lakesentry.io and click Start Free.
2. Enter your email address.
3. Check your inbox for a magic link and click it to verify your account.
4. You’ll land on the LakeSentry setup screen.

Tip

The Free plan includes unlimited workspaces, 3 months of history, and cost visibility for 1 user — no credit card required.

Step 2: Connect your Databricks workspace

This step creates the link between LakeSentry and your Databricks environment. LakeSentry uses a Direct Connection — it connects to your workspace and extracts system table data automatically.

1. From the setup screen, choose Direct Connection as the connection method.
2. Choose your authentication method:
   • Service Principal (OAuth) — Recommended. Uses a dedicated app identity with no user dependency.
   • Personal Access Token — Fastest to set up. Uses a token tied to a user account.
3. Enter your Workspace URL (e.g., https://adb-xxx.azuredatabricks.net). LakeSentry auto-detects the cloud provider and region from this URL.
4. Provide your credentials:
   • For Service Principal: enter the Client ID and Client Secret.
   • For Personal Access Token: enter your Databricks token (starts with dapi).
5. Click Validate Credentials. LakeSentry verifies the connection, checks for SQL warehouses, and confirms system table access.
6. Once validated, click Connect Workspace.

LakeSentry immediately begins extracting data from your Databricks system tables.

For detailed instructions on creating the service principal and configuring permissions, see Connecting Your Databricks Account.

Step 3: Add additional regions (if needed)

Databricks system tables are regional — a connector in one region can’t read data from another. If you run workspaces across multiple regions, add a connector for each region.

1. Go to Settings > Connector and click Add Connector.
2. Enter the Workspace URL for a workspace in the additional region.
3. Provide credentials (the same service principal can be used across regions).
4. LakeSentry auto-detects the region and creates the connector.

Note

The first connector you add becomes the primary connector for global tables. You can change this later in the connector settings.

Step 4: Wait for initial data

After LakeSentry starts its first extraction, the onboarding progress screen shows four pipeline stages:

1. Data Extraction — LakeSentry queries your Databricks system tables.
2. Ledger Transform — Raw data is transformed into the cost ledger.
3. Metrics — Cost metrics are computed and materialized.
4. Insights — Automated cost insights are generated.

You can check the status of each connector in Settings > Connector. Each connector shows its sync status:

Status	Meaning
Synced (green)	Data was received and the connector is active
Error (red)	Validation or connection errors
Awaiting data (gray)	No data received yet (setup in progress)

Step 5: Explore the dashboard

Once data starts flowing, head to the Overview page to see:

• Cost trends — Total spend over time, broken down by compute type
• Top spenders — The workspaces, jobs, and users driving the most cost
• Anomalies — Unusual cost spikes flagged automatically
• Insights — Actionable findings about waste and optimization opportunities

Use the time range selector in the header to adjust the analysis window (Last 24 hours, 7 days, 30 days, 90 days, 1 year, month/year to date, or a custom range). Apply workspace, organization, or tag filters to focus on specific areas.

For a detailed walkthrough of the dashboard, see Understanding the Dashboard.

Step 6: Investigate your first cost spike

Here’s a typical first investigation:

1. On the Overview page, spot a cost anomaly or top spender.
2. Click through to see the cost breakdown — which workspaces, compute types, and users contributed.
3. Drill into a specific work unit (job, pipeline, SQL query, notebook, or serving endpoint) to see per-run cost history.
4. Check the Insights page for LakeSentry’s automated findings — it may have already flagged the issue.

What’s next

• Connecting Your Databricks Account — Detailed setup for service principals and permissions
• Understanding the Dashboard — Full walkthrough of the Overview page
• Cost Attribution — Learn how costs are assigned to teams
• Insights & Actions — How to act on optimization findings