From 0395c47193b3bbf6b4f060f28c9f632580313a35 Mon Sep 17 00:00:00 2001 From: Yorick Peterse Date: Mon, 10 Jul 2017 17:43:57 +0200 Subject: Migrate events into a new format 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. --- lib/api/entities.rb | 12 +- lib/api/v3/entities.rb | 12 +- .../migrate_events_to_push_event_payloads.rb | 176 +++++++++++++++++++++ lib/gitlab/import_export/import_export.yml | 26 ++- 4 files changed, 216 insertions(+), 10 deletions(-) create mode 100644 lib/gitlab/background_migration/migrate_events_to_push_event_payloads.rb (limited to 'lib') diff --git a/lib/api/entities.rb b/lib/api/entities.rb index 3bb1910a441..18cd604a216 100644 --- a/lib/api/entities.rb +++ b/lib/api/entities.rb @@ -497,14 +497,24 @@ module API expose :author, using: Entities::UserBasic end + class PushEventPayload < Grape::Entity + expose :commit_count, :action, :ref_type, :commit_from, :commit_to + expose :ref, :commit_title + end + class Event < Grape::Entity expose :title, :project_id, :action_name expose :target_id, :target_iid, :target_type, :author_id - expose :data, :target_title + expose :target_title expose :created_at expose :note, using: Entities::Note, if: ->(event, options) { event.note? } expose :author, using: Entities::UserBasic, if: ->(event, options) { event.author } + expose :push_event_payload, + as: :push_data, + using: PushEventPayload, + if: -> (event, _) { event.push? } + expose :author_username do |event, options| event.author&.username end diff --git a/lib/api/v3/entities.rb b/lib/api/v3/entities.rb index 4a2e9c9cbb0..a9a35f2a4bd 100644 --- a/lib/api/v3/entities.rb +++ b/lib/api/v3/entities.rb @@ -25,14 +25,24 @@ module API expose(:downvote?) { |note| false } end + class PushEventPayload < Grape::Entity + expose :commit_count, :action, :ref_type, :commit_from, :commit_to + expose :ref, :commit_title + end + class Event < Grape::Entity expose :title, :project_id, :action_name expose :target_id, :target_type, :author_id - expose :data, :target_title + expose :target_title expose :created_at expose :note, using: Entities::Note, if: ->(event, options) { event.note? } expose :author, using: ::API::Entities::UserBasic, if: ->(event, options) { event.author } + expose :push_event_payload, + as: :push_data, + using: PushEventPayload, + if: -> (event, _) { event.push? } + expose :author_username do |event, options| event.author&.username end diff --git a/lib/gitlab/background_migration/migrate_events_to_push_event_payloads.rb b/lib/gitlab/background_migration/migrate_events_to_push_event_payloads.rb new file mode 100644 index 00000000000..432f7c3e706 --- /dev/null +++ b/lib/gitlab/background_migration/migrate_events_to_push_event_payloads.rb @@ -0,0 +1,176 @@ +module Gitlab + module BackgroundMigration + # Class that migrates events for the new push event payloads setup. All + # events are copied to a shadow table, and push events will also have a row + # created in the push_event_payloads table. + class MigrateEventsToPushEventPayloads + class Event < ActiveRecord::Base + self.table_name = 'events' + + serialize :data + + BLANK_REF = ('0' * 40).freeze + TAG_REF_PREFIX = 'refs/tags/'.freeze + MAX_INDEX = 69 + PUSHED = 5 + + def push_event? + action == PUSHED && data.present? + end + + def commit_title + commit = commits.last + + return nil unless commit && commit[:message] + + index = commit[:message].index("\n") + message = index ? commit[:message][0..index] : commit[:message] + + message.strip.truncate(70) + end + + def commit_from_sha + if create? + nil + else + data[:before] + end + end + + def commit_to_sha + if remove? + nil + else + data[:after] + end + end + + def data + super || {} + end + + def commits + data[:commits] || [] + end + + def commit_count + data[:total_commits_count] || 0 + end + + def ref + data[:ref] + end + + def trimmed_ref_name + if ref_type == :tag + ref[10..-1] + else + ref[11..-1] + end + end + + def create? + data[:before] == BLANK_REF + end + + def remove? + data[:after] == BLANK_REF + end + + def push_action + if create? + :created + elsif remove? + :removed + else + :pushed + end + end + + def ref_type + if ref.start_with?(TAG_REF_PREFIX) + :tag + else + :branch + end + end + end + + class EventForMigration < ActiveRecord::Base + self.table_name = 'events_for_migration' + end + + class PushEventPayload < ActiveRecord::Base + self.table_name = 'push_event_payloads' + + enum action: { + created: 0, + removed: 1, + pushed: 2 + } + + enum ref_type: { + branch: 0, + tag: 1 + } + end + + # start_id - The start ID of the range of events to process + # end_id - The end ID of the range to process. + def perform(start_id, end_id) + return unless migrate? + + find_events(start_id, end_id).each { |event| process_event(event) } + end + + def process_event(event) + replicate_event(event) + create_push_event_payload(event) if event.push_event? + end + + def replicate_event(event) + new_attributes = event.attributes + .with_indifferent_access.except(:title, :data) + + EventForMigration.create!(new_attributes) + rescue ActiveRecord::InvalidForeignKey + # A foreign key error means the associated event was removed. In this + # case we'll just skip migrating the event. + end + + def create_push_event_payload(event) + commit_from = pack(event.commit_from_sha) + commit_to = pack(event.commit_to_sha) + + PushEventPayload.create!( + event_id: event.id, + commit_count: event.commit_count, + ref_type: event.ref_type, + action: event.push_action, + commit_from: commit_from, + commit_to: commit_to, + ref: event.trimmed_ref_name, + commit_title: event.commit_title + ) + rescue ActiveRecord::InvalidForeignKey + # A foreign key error means the associated event was removed. In this + # case we'll just skip migrating the event. + end + + def find_events(start_id, end_id) + Event + .where('NOT EXISTS (SELECT true FROM events_for_migration WHERE events_for_migration.id = events.id)') + .where(id: start_id..end_id) + end + + def migrate? + Event.table_exists? && PushEventPayload.table_exists? && + EventForMigration.table_exists? + end + + def pack(value) + value ? [value].pack('H*') : nil + end + end + end +end diff --git a/lib/gitlab/import_export/import_export.yml b/lib/gitlab/import_export/import_export.yml index c5c05bfe2fb..9d9ebcb389a 100644 --- a/lib/gitlab/import_export/import_export.yml +++ b/lib/gitlab/import_export/import_export.yml @@ -3,18 +3,22 @@ project_tree: - labels: :priorities - milestones: - - :events + - events: + - :push_event_payload - issues: - - :events + - events: + - :push_event_payload - :timelogs - notes: - :author - - :events + - events: + - :push_event_payload - label_links: - label: :priorities - milestone: - - :events + - events: + - :push_event_payload - snippets: - :award_emoji - notes: @@ -25,21 +29,25 @@ project_tree: - merge_requests: - notes: - :author - - :events + - events: + - :push_event_payload - merge_request_diff: - :merge_request_diff_commits - :merge_request_diff_files - - :events + - events: + - :push_event_payload - :timelogs - label_links: - label: :priorities - milestone: - - :events + - events: + - :push_event_payload - pipelines: - notes: - :author - - :events + - events: + - :push_event_payload - :stages - :statuses - :triggers @@ -107,6 +115,8 @@ excluded_attributes: statuses: - :trace - :token + push_event_payload: + - :event_id methods: labels: -- cgit v1.2.1 From aac1de46c9be659b74da12f704412f38292974db Mon Sep 17 00:00:00 2001 From: Yorick Peterse Date: Thu, 27 Jul 2017 19:42:15 +0200 Subject: Use a specialized class for querying events 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. --- lib/gitlab/database.rb | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'lib') diff --git a/lib/gitlab/database.rb b/lib/gitlab/database.rb index d7dab584a44..e001d25e7b7 100644 --- a/lib/gitlab/database.rb +++ b/lib/gitlab/database.rb @@ -25,6 +25,10 @@ module Gitlab database_version.match(/\A(?:PostgreSQL |)([^\s]+).*\z/)[1] end + def self.join_lateral_supported? + postgresql? && version.to_f >= 9.3 + end + def self.nulls_last_order(field, direction = 'ASC') order = "#{field} #{direction}" -- cgit v1.2.1