Check configurations

Related documentation

• Checks
• Check properties
• dbt check command

Available configurations

Project YAML file

dbt_project.yml

checks:
  <resource-path>:
    +severity: error | warn
    +enabled: true | false
    +selection_filter_on: column_name | [column_names] | none
    +tags: string | [string]
    +meta: {dictionary}

Properties YAML file

checks/_checks.yml

version: 2

checks:
  - name: check-name
    config:
      severity: error | warn
      enabled: true | false
      selection_filter_on: column_name | [column_names] | none
      tags: string | [string]
      meta: {dictionary}

SQL config

checks/<check_name>.sql

{{ config(
    severity = "error" | "warn",
    enabled = true | false,
    selection_filter_on = "column_name" | ["column_names"] | "none",
    tags = ["string"],
    meta = {"key": "value"}
) }}

select ...
from {{ info_schema('models') }}
where ...

Examples

The following examples show common ways to configure checks.

Warn on check failure

You can use severity: warn when rolling out a new rule gradually. Issues are logged but the build does not fail.

checks/_public_models_have_owners.yml

checks:
  - name: public_models_have_owners
    config:
      severity: warn  # default is error

Filter by a specific column

The edges table has no unique_id column, so checks that query it won't return one. When you use --select, dbt looks for a unique_id column to scope results and finds none, so the check runs against the whole project regardless of the selector. Set selection_filter_on to the columns that contain resource IDs so --select scopes rows by those columns. For example, the multiple_sources_joined check aggregates by child_unique_id, so only that column needs to be set:

checks/_multiple_sources_joined.yml

checks:
  - name: multiple_sources_joined
    config:
      selection_filter_on: child_unique_id

Was this page helpful?

Privacy policyCreate a GitHub issue

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.