Skip to main content

Freshness JSON file Beta

This file contains freshness results for all sources and models with freshness configured. dbt writes it to target/freshness.json after every dbt freshness invocation.

Each entry includes a resource_type field that identifies whether a resource is a source or a model.

Top-level keys​

  • metadata: Invocation metadata with the following fields:
    • dbt_schema_version: Schema URL for this artifact (always https://schemas.getdbt.com/dbt/freshness/v0.json).
    • dbt_version: dbt version that produced the file.
    • generated_at: Timestamp when the file was written.
    • invocation_id: Unique identifier for this invocation.
    • invocation_started_at: Timestamp when the invocation started. Omitted if not available.
    • env: Map of DBT_ENV_CUSTOM_ENV_* environment variables, if set.
  • results: Array of freshness results.

Each entry in results is a dictionary with the following keys:

  • unique_id: Unique node identifier, mapping results to nodes or sources in the manifest.
  • resource_type: Either model or source.
  • max_loaded_at: Max value of the loaded-at timestamp when queried.
  • snapshotted_at: Timestamp when the check ran.
  • max_loaded_at_time_ago_in_s: Interval between max_loaded_at and snapshotted_at in seconds.
  • status: Freshness status: Pass, Warn, or Error. If the freshness query fails, dbt logs an error but omits the node from the file rather than writing it with a failed status.
  • criteria: The freshness configuration for this node, including warn_after, error_after, filter, loaded_at_field, and loaded_at_query.
  • thread_id: Identifier for the thread that ran this check.

Was this page helpful?

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

0
Loading