1. 16 Dec, 2018 1 commit
  2. 10 Dec, 2018 1 commit
    • Gabriel Mazetto's avatar
      Reduce N+1 from Activity Dashboard and Banzai · 4f5abe43
      Gabriel Mazetto authored
      There is a combination of few strategies implemented here:
      
      1. Few relations were eager loaded
      2. Changed few polymorphic routes to specific ones so we don't have to
         use `#becomes(Namespace)` which doesn't preserve association cache
      4f5abe43
  3. 21 Sep, 2018 2 commits
  4. 17 Sep, 2018 1 commit
    • Yorick Peterse's avatar
      Added FromUnion to easily select from a UNION · 8a72f5c4
      Yorick Peterse authored
      This commit adds the module `FromUnion`, which provides the class method
      `from_union`. This simplifies the process of selecting data from the
      result of a UNION, and reduces the likelihood of making mistakes. As a
      result, instead of this:
      
          union = Gitlab::SQL::Union.new([foo, bar])
      
          Foo.from("(#{union.to_sql}) #{Foo.table_name}")
      
      We can now write this instead:
      
          Foo.from_union([foo, bar])
      
      This commit also includes some changes to make this new setup work
      properly. For example, a bug in Rails 4
      (https://github.com/rails/rails/issues/24193) would break the use of
      `from("sub-query-here").includes(:relation)` in certain cases. There was
      also a CI query which appeared to repeat a lot of conditions from an
      outer query on an inner query, which isn't necessary.
      
      Finally, we include a RuboCop cop to ensure developers use this new
      module, instead of using Gitlab::SQL::Union directly.
      
      Fixes https://gitlab.com/gitlab-org/gitlab-ce/issues/51307
      8a72f5c4
  5. 07 Sep, 2018 1 commit
  6. 27 Jul, 2018 1 commit
  7. 29 May, 2018 1 commit
  8. 11 Apr, 2018 1 commit
    • blackst0ne's avatar
      [Rails5] Update Event#subclass_from_attributes method · c57cddf5
      blackst0ne authored
      In Rails 5.0 the `ActiveRecord::Inheritance::subclass_from_attributes`
      method was updated.
      Now it calls the `find_sti_class` method [1] which is overriden in the `Event`
      model and returns needed class (`Event` vs `PushEvent`).
      
      [1]: https://github.com/rails/rails/blob/5-0-stable/activerecord/lib/active_record/inheritance.rb#L209
      
      This commit fixes the errors like
      ```
      143) User#contributed_projects doesn't include IDs for unrelated projects
            Failure/Error: action = attrs.with_indifferent_access[inheritance_column].to_i
      
            NoMethodError:
              undefined method `with_indifferent_access' for nil:NilClass
            # ./app/models/event.rb:118:in `subclass_from_attributes'
      ```
      which are raised on the `RAILS5=1 rspec ...` command.
      c57cddf5
  9. 23 Mar, 2018 1 commit
    • Andreas Brandl's avatar
      Remove N+1 query for Noteable association. · 783868e9
      Andreas Brandl authored
      For each event with `Event#target_type` of "Note",
      we had a query to load the associated instance `noteable`.
      
      For example, if `Note` was related to an issue, we'd load each `Issue`
      with its own query (N+1 problem).
      
      Closes #43150.
      783868e9
  10. 06 Mar, 2018 4 commits
  11. 05 Mar, 2018 1 commit
  12. 13 Feb, 2018 1 commit
  13. 04 Jan, 2018 1 commit
    • Yorick Peterse's avatar
      Eager load event target authors whenever possible · dac51ace
      Yorick Peterse authored
      This ensures that the "author" association of an event's "target"
      association is eager loaded whenever the "target" association defines an
      "author" association. This in turn solves the N+1 query problem we first
      tried to solve in
      https://gitlab.com/gitlab-org/gitlab-ce/merge_requests/15788 but caused
      problems when displaying milestones as those don't define an "author"
      association.
      
      The approach in this commit does mean that the authors are _always_
      eager loaded since this takes place in the "belongs_to" block. This
      however shouldn't pose too much of a problem, and as far as I can tell
      there's no real way around this unfortunately.
      dac51ace
  14. 22 Dec, 2017 1 commit
  15. 08 Dec, 2017 1 commit
    • Yorick Peterse's avatar
      Fix N+1 query when displaying events · c52a36e8
      Yorick Peterse authored
      When displaying events we would load the target of those events, then
      render the entire data using our Markdown pipeline. This pipeline would
      eventually request the author of every target, leading to an additional
      query being executed for every target to get the author.
      
      To fix this we now eager load the author of the event's target. In my
      local environment this reduces the number of queries to display a
      project's Atom feed from 40 to 24 queries.
      
      See https://gitlab.com/gitlab-org/gitlab-ce/issues/36878 for more
      information.
      c52a36e8
  16. 20 Sep, 2017 1 commit
    • Yorick Peterse's avatar
      Remove redundant WHERE from event queries · 9a3e4b8d
      Yorick Peterse authored
      The default scope in Event would add a "WHERE author_id IS NOT NULL"
      clause to every query. Now that "events.author_id" has a NOT NULL clause
      set this filter is redundant and we can safely remove it.
      
      In this commit we also add a validation on events.author_id (previously
      only defined in PushEvent) just in case something tries to create data
      without an author ID. This way we can present a nicer error message
      compared to PostgreSQL's foreign key error messages.
      
      Fixes https://gitlab.com/gitlab-org/gitlab-ce/issues/38129
      9a3e4b8d
  17. 12 Sep, 2017 1 commit
  18. 08 Sep, 2017 1 commit
    • Yorick Peterse's avatar
      Rework how recent push events are retrieved · 83355336
      Yorick Peterse authored
      Whenever you push to a branch GitLab will show a button to create a
      merge request (should one not exist already). The underlying code to
      display this data was quite inefficient. For example, it involved
      multiple slow queries just to figure out what the most recent push event
      was.
      
      This commit changes the way this data is retrieved so it's much faster.
      This is achieved by caching the ID of the last push event on every push,
      which is then retrieved when loading certain pages. Database queries are
      only executed if necessary and the cached data is removed automatically
      once a merge request has been created, or 2 hours after being stored.
      
      A trade-off of this approach is that we _only_ track the last event.
      Previously if you were to push to branch A and B then create a merge
      request for branch B we'd still show the widget for branch A. As of this
      commit this is no longer the case, instead we will only show the widget
      for the branch you pushed to most recently. Once a merge request exists
      the widget is no longer displayed. Alternative solutions are either too
      complex and/or too slow, hence the decision was made to settle for this
      trade-off.
      
      Performance Impact
      ------------------
      
      In the best case scenario (= a user didn't push anything for more than 2
      hours) we perform a single Redis GET per page. Should there be cached
      data we will run a single (and lightweight) SQL query to get the
      event data from the database. If a merge request already exists we will
      run an additional DEL to remove the cache key.
      
      The difference in response timings can vary a bit per project. On
      GitLab.com the 99th percentile of time spent in User#recent_push hovers
      between 100 milliseconds and 1 second, while the mean hovers around 50
      milliseconds. With the changes in this MR the expected time spent in
      User#recent_push is expected to be reduced down to just a few
      milliseconds.
      
      Fixes https://gitlab.com/gitlab-org/gitlab-ce/issues/35990
      83355336
  19. 06 Sep, 2017 1 commit
    • Yorick Peterse's avatar
      Finish migration to the new events setup · 235b105c
      Yorick Peterse authored
      This finishes the procedure for migrating events from the old format
      into the new format. Code no longer uses the old setup and the database
      tables used during the migration process are swapped, with the old table
      being dropped.
      
      While the database migration can be reversed this will 1) take a lot of
      time as data has to be coped around 2) won't restore data in the
      "events.data" column as we have no way of restoring this.
      
      Fixes https://gitlab.com/gitlab-org/gitlab-ce/issues/37241
      235b105c
  20. 22 Aug, 2017 1 commit
  21. 21 Aug, 2017 1 commit
  22. 10 Aug, 2017 2 commits
    • Yorick Peterse's avatar
      Use a specialized class for querying events · aac1de46
      Yorick Peterse authored
      This changes various controllers to use the new EventCollection class
      for retrieving events. This class uses a JOIN LATERAL query on
      PostgreSQL to retrieve queries in a more efficient way, while falling
      back to a simpler / less efficient query for MySQL.
      
      The EventCollection class also includes a limit on the number of events
      to display to prevent malicious users from cycling through all events,
      as doing so could put a lot of pressure on the database.
      
      JOIN LATERAL is only supported on PostgreSQL starting with version 9.3.0
      and as such this optimisation is only used when using PostgreSQL 9.3 or
      newer.
      aac1de46
    • Yorick Peterse's avatar
      Migrate events into a new format · 0395c471
      Yorick Peterse authored
      This commit migrates events data in such a way that push events are
      stored much more efficiently. This is done by creating a shadow table
      called "events_for_migration", and a table called "push_event_payloads"
      which is used for storing push data of push events. The background
      migration in this commit will copy events from the "events" table into
      the "events_for_migration" table, push events in will also have a row
      created in "push_event_payloads".
      
      This approach allows us to reclaim space in the next release by simply
      swapping the "events" and "events_for_migration" tables, then dropping
      the old events (now "events_for_migration") table.
      
      The new table structure is also optimised for storage space, and does
      not include the unused "title" column nor the "data" column (since this
      data is moved to "push_event_payloads").
      
      == Newly Created Events
      
      Newly created events are inserted into both "events" and
      "events_for_migration", both using the exact same primary key value. The
      table "push_event_payloads" in turn has a foreign key to the _shadow_
      table. This removes the need for recreating and validating the foreign
      key after swapping the tables. Since the shadow table also has a foreign
      key to "projects.id" we also don't have to worry about orphaned rows.
      
      This approach however does require some additional storage as we're
      duplicating a portion of the events data for at least 1 release. The
      exact amount is hard to estimate, but for GitLab.com this is expected to
      be between 10 and 20 GB at most. The background migration in this commit
      deliberately does _not_ update the "events" table as doing so would put
      a lot of pressure on PostgreSQL's auto vacuuming system.
      
      == Supporting Both Old And New Events
      
      Application code has also been adjusted to support push events using
      both the old and new data formats. This is done by creating a PushEvent
      class which extends the regular Event class. Using Rails' Single Table
      Inheritance system we can ensure the right class is used for the right
      data, which in this case is based on the value of `events.action`. To
      support displaying old and new data at the same time the PushEvent class
      re-defines a few methods of the Event class, falling back to their
      original implementations for push events in the old format.
      
      Once all existing events have been migrated the various push event
      related methods can be removed from the Event model, and the calls to
      `super` can be removed from the methods in the PushEvent model.
      
      The UI and event atom feed have also been slightly changed to better
      handle this new setup, fortunately only a few changes were necessary to
      make this work.
      
      == API Changes
      
      The API only displays push data of events in the new format. Supporting
      both formats in the API is a bit more difficult compared to the UI.
      Since the old push data was not really well documented (apart from one
      example that used an incorrect "action" nmae) I decided that supporting
      both was not worth the effort, especially since events will be migrated
      in a few days _and_ new events are created in the correct format.
      0395c471
  23. 06 Jul, 2017 1 commit
  24. 21 Jun, 2017 1 commit
  25. 07 Jun, 2017 1 commit
  26. 06 Jun, 2017 1 commit
    • Mark Fletcher's avatar
      Introduce an Events API · ad3e180e
      Mark Fletcher authored
      * Meld the following disparate endpoints:
       * `/projects/:id/events`
       * `/events`
       * `/users/:id/events`
      + Add result filtering to the above endpoints:
       * action
       * target_type
       * before and after dates
      ad3e180e
  27. 31 May, 2017 1 commit
  28. 04 May, 2017 1 commit
  29. 26 Apr, 2017 1 commit
  30. 21 Mar, 2017 1 commit
  31. 25 Feb, 2017 1 commit
  32. 23 Feb, 2017 3 commits
  33. 13 Feb, 2017 1 commit