MetricSign
EN|NLRequest Access
Medium severitydbt

Power BI Refresh Error:
Package Install Error

What does this error mean?

The `dbt deps` command failed to install one or more packages from packages.yml. This blocks all subsequent dbt commands because package macros and models are unavailable. Common in dbt Cloud when Hub packages or Git-based packages are unreachable.

Common causes

  • 1A package version specified in packages.yml does not exist on dbt Hub or has been yanked
  • 2A Git-based package URL is unreachable from the dbt Cloud environment (private repo, expired token)
  • 3Package version constraints conflict — two packages require incompatible versions of a shared dependency
  • 4The dbt Hub registry is temporarily unavailable (rare, but transient failures occur)
  • 5The dbt-core version in the Cloud environment is incompatible with the declared package versions

How to fix it

  1. 1Run `dbt deps` locally and compare the output to the Cloud job failure — if it works locally, the issue is network access in dbt Cloud.
  2. 2Check dbt Hub (hub.getdbt.com) to verify the package version you specified exists and is compatible with your dbt-core version.
  3. 3For Git-based packages, verify the branch/tag/commit reference is still valid and the repository is accessible with the configured credentials.
  4. 4If two packages conflict, check their changelogs to find a compatible version combination or pin one to an older version.
  5. 5Try pinning the package to an exact version rather than a range: `version: 1.2.3` instead of `version: ['>=1.0.0', '<2.0.0']`.

Frequently asked questions

Should I run `dbt deps` in every dbt Cloud job?

Yes — always include `dbt deps` as the first step in your Cloud job to ensure packages match packages.yml. Skipping it can result in stale macros from a previous package version.

How do I use a private Git repository as a dbt package?

In packages.yml, add the package with `git: <repo_url>` and `revision: <branch_or_tag>`. For private repos, configure a GitHub/GitLab integration or personal access token in dbt Cloud settings — without it, `dbt deps` fails with 401 or 403.

Can version conflicts between packages cause models to behave incorrectly without an error?

Yes — if two packages declare macros with the same name, the last installed wins silently. After resolving conflicts, run `dbt clean && dbt deps` and verify macro behavior.

Official documentation: https://docs.getdbt.com/docs/build/packages

Other dbt errors