The dbt process does not have write permission to the target/ directory. The disk or ephemeral storage on the CI runner or dbt Cloud instance is full. The target directory was deleted or is on a read-only filesystem (e.g., a mounted volume in Docker). A concurrent dbt process is writing to the same target directory simultaneously

How do I fix IOError?

Step 1: Check disk usage on the runner: df -h. Clean up large files or expand the storage volume.. Step 2: Verify the dbt process user has write access to the target/ directory: ls -la dbt_project/target/.. Step 3: In Docker or Kubernetes, ensure the target directory is mounted with read-write permissions.. Step 4: If running concurrent dbt processes, use separate --target-path values to avoid conflicts.. Step 5: In dbt Cloud, contact support if artifact writes fail on the managed infrastructure.

Medium severityexecution

Power BI Refresh Error:
IOError

What does this error mean?

dbt failed to write a run artifact (manifest.json, run_results.json, catalog.json, or sources.json) to the target directory because of a permissions error, a full disk, or a missing directory.

Common causes

1The dbt process does not have write permission to the target/ directory
2The disk or ephemeral storage on the CI runner or dbt Cloud instance is full
3The target directory was deleted or is on a read-only filesystem (e.g., a mounted volume in Docker)
4A concurrent dbt process is writing to the same target directory simultaneously

How to fix it

1Step 1: Check disk usage on the runner: df -h. Clean up large files or expand the storage volume.
2Step 2: Verify the dbt process user has write access to the target/ directory: ls -la dbt_project/target/.
3Step 3: In Docker or Kubernetes, ensure the target directory is mounted with read-write permissions.
4Step 4: If running concurrent dbt processes, use separate --target-path values to avoid conflicts.
5Step 5: In dbt Cloud, contact support if artifact writes fail on the managed infrastructure.

Frequently asked questions

Can I change where dbt writes its artifacts?

Yes — use the --target-path flag or set target-path in dbt_project.yml. Point it to a writable location or a mounted cloud storage path.

Will this error cause my dbt models to fail or just the artifact write?

Models themselves run in the warehouse and are unaffected by local disk issues. The artifact write failure occurs after model execution, so your data is likely correct — but state comparison and documentation will be stale.