Skip to content

Fix ProjectAttributesPipeline re-enqueue in BulkImport

What does this MR do and why?

Add the missing relation name to the ProjectAttributesPipeline pipeline as the name is required by the BulkImports::PipelineWorker to verify if the export file was generated in the source instance

Note: Changelog wasn't added, because the ProjectAttributesPipeline is behind the feature flag bulk_import_projects that is disabled by default.

Context

The BulkImports::PipelineWorker expects that the pipelines that extract data from a file to define a relation name. For the ProjectAttributesPipeline, the relation isn't defined which makes the pipeline to be marked as failed.

Related to: #358745 (closed) and !86597 (merged)

Screenshots or screen recordings

project-attributes

ProjectAttributesPipeline migrates several fields of the projects. The screen recording highlights that the project's merge requests settings are being migrated as they are some settings that are migrated by the pipeline.

How to set up and validate locally

  1. Feature.enable(:bulk_import).
  2. Feature.enable(:bulk_import_projects).
  3. Create a top-level group and a project in the group.
  4. Go to /groups/new#import-group-pane page and enter the instance URL and access token (needs to be api & read_repository scope).
  5. Select the newly created group and click Import.
  6. Wait for Group import to complete and verify the imported group and project data.

MR acceptance checklist

This checklist encourages us to confirm any changes have been analyzed to reduce risks in quality, performance, reliability, security, and maintainability.

Edited by Rodrigo Tomonari

Merge request reports