Skip to main content

Databricks

note

This feature is available as an early‑access capability for Databricks. To enable it for your account, contact Flexera Support. For more information, see Contacting Flexera Support.

Flexera One uses bill data to provide an accurate view of your costs across accounts and services. This data is made available for pre-built and ad-hoc analyses. To gather this cost information, you must perform certain configuration steps and share specific data and credentials with Flexera One.

For Databricks, Flexera One ingests billing, compute, and lakeflow data so you can allocate costs, identify waste, and analyze your Databricks spend in detail.

Overview of Databricks Bill Connect

The Databricks bill connect enables you to ingest your Databricks billing and compute data into Flexera One for cost reporting and analysis.

Flexera One pulls data through Databricks system tables and enriches it to provide detailed insights into your Databricks costs. The Databricks bill connect requires an SQL warehouse and a service principal with access to system tables. The service principal uses the warehouse to pull data for all workspaces in that region. To enable ingestion for additional regions, you must configure a warehouse in at least one workspace per region that the service principal can use and where the service principal has access to system tables.

Getting Started With Databricks Bill Connect

Complete the following steps to connect your Databricks billing and compute data to Flexera One for cost reporting purposes:

  1. Review the Prerequisites
  2. Creating a Service Principal and Granting Workspace and Warehouse Access
  3. Granting System Table and Workspace Discovery Access
  4. Connecting Databricks in Flexera One
  5. Verifying Bill Connect
  6. Viewing Import History to Verify Bill Status
note

You can view your Databricks costs in the built-in Databricks Analyzer and Resource Analyzer dashboards after you complete the bill connect configuration. For more information, see The "Default" Dashboards.

Mapping Flexera One Dimensions to Databricks Billing Columns

The following table describes the Flexera One dimensions and indicates which dimensions map Databricks billing columns:

Flexera One DimensionDatabricksMeaning
Databricks Usage NameValues can be warehouse.name, clusters.name, or pipeline_details.nameThe Databricks compute resource that is being billed, such as a warehouse name, cluster name, job name, pipeline name, or notebook (for serverless).
Databricks Workflow NameValues can be usage_metadata.job_name, the notebook path, or the pipeline name.The workflow that triggers the compute usage.
Billing Account IDusage.account_idA unique identifier for the Databricks account.
Cloud Vendor Accountworkspace_idA unique identifier for the Databricks workspace.
Cloud Vendor Account Nameworkspace_nameA human-readable name for the workspace ID.
CategoryDerived dynamically based on usage.billing_origin_productThe category this usage record belongs to (for example, Compute, Storage, or Network).
Instance Typeusage_metadata.node_typeAn identifier for common hardware configurations for VMs running a customer workload. For example, m4.xlarge (AWS), D2 v3 (Azure), n1-standard-4 (Google Cloud Platform), or Serverless (Databricks).
Line Item TypeUsageIndicates the type of information the line item is for, such as usage, tax, fee, or credit. For Databricks, this value is always Usage.
RegionDerived from usage.sku_name / clusters.aws_attributes.zone_idThe physical region where a resource was consumed.
Resource Typeusage.sku_nameIndicates the type of resource consumed within a service.
Resource IDusage.usage_metadata.*_idThe ID of a specific cloud resource.
Serviceusage.billing_origin_productThe name of the high-level service consumed. For example, AmazonS3 (AWS), Microsoft.Compute (Azure), or Compute Engine (Google Cloud Platform).
Usage Typeusage.usage_typeIndicates information about the kind of usage incurred.
Tagsusage.custom_tags + clusters.tags + jobs.tags + warehouses.tagsTags used for defining custom dimensions.
Usage Amountusage.usage_quantityThe amount of usage generated in this record, in the units specified in Usage Unit (set to 0 if not needed).
Usage Unitusage.usage_unitThe units that the Usage Amount metric is reported in.
Costusage.usage_quantity * list_prices.pricing(usage_unit)The total cost generated in this record, in the currency specified in Currency Code.