What causes DeferralStateManifestNotFound?

The --state path points to a directory that does not contain a manifest.json, or the path itself is incorrect. The production manifest artifact was not downloaded or mounted before the CI/CD run that uses --defer --state. A dbt Cloud job configured to defer to another job's artifact cannot locate the artifact because the referenced job has not run recently or its artifacts were purged

How do I fix DeferralStateManifestNotFound?

Step 1: Verify the --state path exists and contains a manifest.json: ls -la /manifest.json.. Step 2: In CI/CD pipelines, add a step before the dbt run to download the production manifest artifact from dbt Cloud (via the dbt Cloud API artifacts endpoint) or from your artifact storage (S3, GCS).. Step 3: In dbt Cloud, verify that the 'Defer to a previous run state' job configuration points to a job that ran successfully within the artifact retention window.. Step 4: Set the DBT_STATE environment variable to the correct path as an alternative to passing --state on each command.

High severityconfiguration

Power BI Refresh Error:
DeferralStateManifestNotFound

What does this error mean?

A dbt run with --defer or --state failed because the manifest.json file specified via the --state flag does not exist at the provided path, making it impossible to resolve deferred references or compare node states.

Common causes

1The --state path points to a directory that does not contain a manifest.json, or the path itself is incorrect
2The production manifest artifact was not downloaded or mounted before the CI/CD run that uses --defer --state
3A dbt Cloud job configured to defer to another job's artifact cannot locate the artifact because the referenced job has not run recently or its artifacts were purged

How to fix it

1Step 1: Verify the --state path exists and contains a manifest.json: ls -la <state_path>/manifest.json.
2Step 2: In CI/CD pipelines, add a step before the dbt run to download the production manifest artifact from dbt Cloud (via the dbt Cloud API artifacts endpoint) or from your artifact storage (S3, GCS).
3Step 3: In dbt Cloud, verify that the 'Defer to a previous run state' job configuration points to a job that ran successfully within the artifact retention window.
4Step 4: Set the DBT_STATE environment variable to the correct path as an alternative to passing --state on each command.

Frequently asked questions

Does dbt fail hard or warn when the state manifest is missing?

dbt fails hard — it raises a fatal error and exits without running any models if the --state path does not contain a valid manifest.json.

Can I use a partial or older manifest for deferral?

Yes, with caveats. An older manifest works but may defer to stale production nodes. A manifest from a different dbt version may cause a manifest version incompatibility error instead.