Skip to content

Updates for larger datasets

Grant Young requested to merge gy-chromium into main

MR brings in various updates and improvements that we're required for the new Chromium dataset pipeline. The updates are as follows:

  • Adjusted timeouts for Git calls to be a maximum of 10 minutes
  • Added new All Commits API test api_v4_projects_repository_commits_all
  • Changed name of api_v4_projects_repository_files_file.js test to api_v4_projects_repository_files_file_metadata.js to be more reflective
  • Adjusted the following tests to switch to the k6 ramping-arrival-rate executor to ensure RPS rate is not affected by large downloads that these endpoints can have with larger repos - api_v4_projects_repository_files_file_blame, api_v4_projects_repository_files_file_raw.js, git_ls_remote.js, git_pull.js and web_project_files.js.
    • Also ensured discardResponseBodies is set for all to prevent k6 memory exhaustion.
    • Change only impacts larger datasets. Existing datasets see similar results.
  • Added new env var option to see total data downloaded by k6 in results summary - GPT_DATA_RECEIVED
  • Added env var option to change the report name in Slack and Wiki reports to be different from environment name
  • Add config files for initial Chromium project
  • Bumped k6 to v0.50.0
  • Bumped GPT version to 2.15.0

Merge request reports