GitLab CI "This job depends on other jobs with expired/erased artifacts"

By Daniel Zoghalchali·Latchkey

When a job uses needs (or dependencies) to pull artifacts from an earlier job, those artifacts must still exist. If their expiry passed or they were manually erased, the download fails.

What this error means

The job shows a message that it depends on other jobs with expired or erased artifacts, and refuses to start or fails immediately when retried much later.

This job depends on other jobs with expired/erased artifacts:
build. Please refer to the documentation on artifact expiry.

Diagnose it: which rule matched, and on which runner?

GitLab evaluates rules: top to bottom and the first match wins, including one that sets when: never. A job that does not run, or runs when you did not expect it to, is nearly always matching an earlier rule than the one you are reading.

# validate the definition against the project
curl -s --header "PRIVATE-TOKEN: $TOKEN" \
  "https://gitlab.com/api/v4/projects/$CI_PROJECT_ID/ci/lint" \
  --data-urlencode "content=$(cat .gitlab-ci.yml)"

# what the job actually sees
script:
  - env | grep -E "^CI_(PIPELINE_SOURCE|COMMIT_REF_NAME|RUNNER)" | sort

Common causes

Artifacts expired

The upstream job artifacts:expire_in elapsed before this job retried, so they were cleaned up.

Artifacts manually erased

A maintainer erased the upstream job, deleting its artifacts.

Retrying an old pipeline

Re-running a downstream job long after the pipeline finished finds the inputs already gone.

How to fix it

Re-run the upstream job to regenerate artifacts

Retry from the producing job (or the whole pipeline) so artifacts are recreated before the consumer runs.

1. Open the pipeline and retry the upstream job that produces the artifacts.
2. Wait for it to upload artifacts, then retry the dependent job.
3. If the pipeline is old, run a fresh pipeline instead.

Extend artifact expiry for long pipelines

Increase expire_in on producers whose outputs feed late-running consumers.

build:
  script: make build
  artifacts:
    paths: [dist/]
    expire_in: 1 week

How to prevent it

• Set artifacts:expire_in long enough to cover the slowest downstream consumer.
• Avoid retrying single downstream jobs in very old pipelines.
• Use keep:latest on critical artifacts where supported.

Frequently asked questions

Related guides

References

• GitHub Actions documentation