Skip to content

Job does not fail when docker image does not exist #301

@valkum

Description

@valkum

We had a job that showed success even though the job never run due to:

Error response from daemon: manifest unknown
Exited with 1
Retrying 2 more times...
Error response from daemon: manifest unknown
Exited with 1
Retrying 1 more times...
Error response from daemon: manifest unknown
Exited with 1
Failed 3 retries
!!! :docker: Pull failed.
steps:
  - label: "Run Daily Job"
    if: build.source == "schedule" || build.source == "ui"
    concurrency_group: ...
    concurrency: 1
    plugins:
      - gcp-workload-identity-federation#v1.5.0:
         ...
      - docker#v5.11.0:
          image: IMAGE THAT DOES NOT EXIST
          always-pull: true
          propagate-environment: true
          propagate-gcp-auth-tokens: true
          tty: false
          environment:
            - ENV.....
          mounts:
            - type: volume
              source: cache
              target: /.cache
    command: |
      some command

I would assume this should fail.
Related issue #279

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions