BigW Consortium Gitlab

  1. 11 Aug, 2017 1 commit
  2. 01 Aug, 2017 1 commit
  3. 20 Jul, 2017 1 commit
  4. 18 Jul, 2017 1 commit
  5. 12 Jun, 2017 1 commit
  6. 08 Jun, 2017 3 commits
    • Bring in security changes from the 9.2.5 release · 565ead61
      DJ Mountney authored
      Ran:
       - git format-patch v9.2.2..v9.2.5 --stdout > patchfile.patch
       - git checkout -b 9-2-5-security-patch origin/v9.2.2
       - git apply patchfile.patch
       - git commit
       - [Got the sha ref for the commit]
       - git checkout -b upstream-9-2-security master
       - git cherry-pick <SHA of the patchfile commit>
       - [Resolved conflicts]
       - git cherry-pick --continue
    • Bring in security changes from the 9.2.5 release · 1d1363e2
      DJ Mountney authored
      Ran:
       - git format-patch v9.2.2..v9.2.5 --stdout > patchfile.patch
       - git checkout -b 9-2-5-security-patch origin/v9.2.2
       - git apply patchfile.patch
       - git commit
       - [Got the sha ref for the commit]
       - git checkout -b upstream-9-2-security master
       - git cherry-pick <SHA of the patchfile commit>
       - [Resolved conflicts]
       - git cherry-pick --continue
  7. 07 Jun, 2017 1 commit
  8. 06 Jun, 2017 1 commit
    • Set artifact working directory to be in the destination store to prevent unnecessary I/O · 8a417f5a
      Stan Hu authored
      Similar to #33218, build artifacts were being uploaded into a CarrierWave
      temporary directory in the Rails root directory before moved to their
      final destination, which could cause a copy across filesystems. This
      merge request refactors the work in !11866 so that any uploader can
      just override `work_dir` to change the default implementation.
      
      Closes #33274
  9. 05 Jun, 2017 1 commit
    • Fix LFS timeouts when trying to save large files · 37dd0737
      Stan Hu authored
      The following was happening:
      
      1. Workhorse stores an LFS file in /var/opt/gitlab/gitlab-rails/shared/lfs-objects
      2. CarrierWave then renames the file to a temporary directory (e.g. /opt/gitlab/embedded/service/gitlab-rails/tmp)
      3. CarrierWave then renames the file to its final location (e.g. /var/opt/gitlab/gitlab-rails/shared/lfs-objects)
      
      When the LFS upload path was on a different filesystem than the Rails
      installation, step 2 could take a longer than 10 seconds, at which point Workhorse
      would time out with "badgateway: failed after 10s: context canceled".
      
      This change makes the work path in the same root as the LFS storage path, preventing moves
      across filesystems.
      
      Closes #33218
  10. 01 Jun, 2017 2 commits
  11. 31 May, 2017 1 commit
  12. 02 May, 2017 1 commit
  13. 09 Apr, 2017 1 commit
  14. 06 Mar, 2017 2 commits
  15. 24 Feb, 2017 1 commit
    • Minor refactoring of Uploaders · a8c62dfe
      Robert Speicher authored
      - Moves a duplicate `file_storage?` definition into the common
        `GitlabUploader` ancestor.
      - Get the `uploads` base directory from a class method rather than
        hard-coding it where it's needed. This will be used in a subsequent MR
        to store Uploads in the database.
      - Improves the specs for uploaders.
  16. 23 Feb, 2017 1 commit
  17. 15 Feb, 2017 1 commit
  18. 03 Jan, 2017 1 commit
  19. 15 Dec, 2016 1 commit
  20. 08 Dec, 2016 1 commit
  21. 23 Nov, 2016 1 commit
    • Remove event caching code · 5371da34
      Yorick Peterse authored
      Flushing the events cache worked by updating a recent number of rows in
      the "events" table. This has the result that on PostgreSQL a lot of dead
      tuples are produced on a regular basis. This in turn means that
      PostgreSQL will spend considerable amounts of time vacuuming this table.
      This in turn can lead to an increase of database load.
      
      For GitLab.com we measured the impact of not using events caching and
      found no measurable increase in response timings. Meanwhile not flushing
      the events cache lead to the "events" table having no more dead tuples
      as now rows are only inserted into this table.
      
      As a result of this we are hereby removing events caching as it does not
      appear to help and only increases database load.
      
      For more information see the following comment:
      https://gitlab.com/gitlab-org/gitlab-ce/merge_requests/6578#note_18864037
  22. 24 Jul, 2016 1 commit
  23. 20 Jul, 2016 1 commit
  24. 19 Jul, 2016 4 commits
  25. 29 Jun, 2016 1 commit
  26. 30 Mar, 2016 4 commits
  27. 15 Mar, 2016 1 commit
  28. 18 Feb, 2016 1 commit
  29. 17 Feb, 2016 1 commit
  30. 16 Feb, 2016 1 commit