Verified Commit 0cc94878 authored by Yorick Peterse's avatar Yorick Peterse
Browse files

Rewrite the GitHub importer from scratch 

Prior to this MR there were two GitHub related importers:

* Github::Import: the main importer used for GitHub projects
* Gitlab::GithubImport: importer that's somewhat confusingly used for
  importing Gitea projects (apparently they have a compatible API)

This MR renames the Gitea importer to Gitlab::LegacyGithubImport and
introduces a new GitHub importer in the Gitlab::GithubImport namespace.
This new GitHub importer uses Sidekiq for importing multiple resources
in parallel, though it also has the ability to import data sequentially
should this be necessary.

The new code is spread across the following directories:

* lib/gitlab/github_import: this directory contains most of the importer
  code such as the classes used for importing resources.
* app/workers/gitlab/github_import: this directory contains the Sidekiq
  workers, most of which simply use the code from the directory above.
* app/workers/concerns/gitlab/github_import: this directory provides a
  few modules that are included in every GitHub importer worker.

== Stages

The import work is divided into separate stages, with each stage
importing a specific set of data. Stages will schedule the work that
needs to be performed, followed by scheduling a job for the
"AdvanceStageWorker" worker. This worker will periodically check if all
work is completed and schedule the next stage if this is the case. If
work is not yet completed this worker will reschedule itself.

Using this approach we don't have to block threads by calling `sleep()`,
as doing so for large projects could block the thread from doing any
work for many hours.

== Retrying Work

Workers will reschedule themselves whenever necessary. For example,
hitting the GitHub API's rate limit will result in jobs rescheduling
themselves. These jobs are not processed until the rate limit has been
reset.

== User Lookups

Part of the importing process involves looking up user details in the
GitHub API so we can map them to GitLab users. The old importer used
an in-memory cache, but this obviously doesn't work when the work is
spread across different threads.

The new importer uses a Redis cache and makes sure we only perform
API/database calls if absolutely necessary.  Frequently used keys are
refreshed, and lookup misses are also cached; removing the need for
performing API/database calls if we know we don't have the data we're
looking for.

== Performance & Models

The new importer in various places uses raw INSERT statements (as
generated by `Gitlab::Database.bulk_insert`) instead of using Rails
models. This allows us to bypass any validations and callbacks,
drastically reducing the number of SQL queries and Gitaly RPC calls
necessary to import projects.

To ensure the code produces valid data the corresponding tests check if
the produced rows are valid according to the model validation rules.
parent 03f8f3d3

app/controllers/import/github_controller.rb

+2 −2
Original line number Diff line number Diff line
@@ -43,7 +43,7 @@ def create 
    @target_namespace = find_or_create_namespace(namespace_path, current_user.namespace_path)



    if can?(current_user, :create_projects, @target_namespace)

      @project = Gitlab::GithubImport::ProjectCreator.new(repo, @project_name, @target_namespace, current_user, access_params, type: provider).execute

      @project = Gitlab::LegacyGithubImport::ProjectCreator.new(repo, @project_name, @target_namespace, current_user, access_params, type: provider).execute

    else

      render 'unauthorized'

    end
@@ -52,7 +52,7 @@ def create 
  private



  def client

    @client ||= Gitlab::GithubImport::Client.new(session[access_token_key], client_options)

    @client ||= Gitlab::LegacyGithubImport::Client.new(session[access_token_key], client_options)

  end



  def verify_import_enabled

app/models/project.rb

+37 −0
Original line number Diff line number Diff line
@@ -354,6 +354,7 @@ def self.with_feature_available_for_user(feature, user) 
  scope :abandoned, -> { where('projects.last_activity_at < ?', 6.months.ago) }



  scope :excluding_project, ->(project) { where.not(id: project) }

  scope :import_started, -> { where(import_status: 'started') }



  state_machine :import_status, initial: :none do

    event :import_schedule do
@@ -1411,6 +1412,31 @@ def after_rename_repo 
    Gitlab::PagesTransfer.new.rename_project(path_before_change, self.path, namespace.full_path)

  end



  def after_import

    repository.after_import

    import_finish

    remove_import_jid

    update_project_counter_caches

  end



  def update_project_counter_caches

    classes = [

      Projects::OpenIssuesCountService,

      Projects::OpenMergeRequestsCountService

    ]



    classes.each do |klass|

      klass.new(self).refresh_cache

    end

  end



  def remove_import_jid

    return unless import_jid



    Gitlab::SidekiqStatus.unset(import_jid)

    update_column(:import_jid, nil)

  end



  def running_or_pending_build_count(force: false)

    Rails.cache.fetch(['projects', id, 'running_or_pending_build_count'], force: force) do

      builds.running_or_pending.count(:all)
@@ -1658,6 +1684,17 @@ def gl_repository(is_wiki:) 
    Gitlab::GlRepository.gl_repository(self, is_wiki)

  end



  # Refreshes the expiration time of the associated import job ID.

  #

  # This method can be used by asynchronous importers to refresh the status,

  # preventing the StuckImportJobsWorker from marking the import as failed.

  def refresh_import_jid_expiration

    return unless import_jid



    Gitlab::SidekiqStatus

      .set(import_jid, StuckImportJobsWorker::IMPORT_JOBS_EXPIRATION)

  end



  private



  def storage

app/services/projects/import_service.rb

+17 −1
Original line number Diff line number Diff line
@@ -4,6 +4,18 @@ class ImportService < BaseService 


    Error = Class.new(StandardError)



    # Returns true if this importer is supposed to perform its work in the

    # background.

    #

    # This method will only return `true` if async importing is explicitly

    # supported by an importer class (`Gitlab::GithubImport::ParallelImporter`

    # for example).

    def async?

      return false unless has_importer?



      importer_class.respond_to?(:async?) && importer_class.async?

    end



    def execute

      add_repository_to_project unless project.gitlab_project_import?
@@ -80,7 +92,11 @@ def has_importer? 
    end



    def importer

      Gitlab::ImportSources.importer(project.import_type).new(project)

      importer_class.new(project)

    end



    def importer_class

      Gitlab::ImportSources.importer(project.import_type)

    end



    def unknown_url?

app/workers/concerns/gitlab/github_import/notify_upon_death.rb

0 → 100644
+28 −0
Original line number Diff line number Diff line 
# frozen_string_literal: true



module Gitlab

  module GithubImport

    # NotifyUponDeath can be included into a GitHub worker class if it should

    # notify any JobWaiter instances upon being moved to the Sidekiq dead queue.

    #

    # Note that this will only notify the waiter upon graceful termination, a

    # SIGKILL will still result in the waiter _not_ being notified.

    #

    # Workers including this module must have jobs passed where the last

    # argument is the key to notify, as a String.

    module NotifyUponDeath

      extend ActiveSupport::Concern



      included do

        # If a job is being exhausted we still want to notify the

        # AdvanceStageWorker. This prevents the entire import from getting stuck

        # just because 1 job threw too many errors.

        sidekiq_retries_exhausted do |job|

          if job.args.length == 3 && (key = job.args.last) && key.is_a?(String)

            JobWaiter.notify(key, job.jid)

          end

        end

      end

    end

  end

end

app/workers/concerns/gitlab/github_import/queue.rb

0 → 100644
+16 −0
Original line number Diff line number Diff line 
module Gitlab

  module GithubImport

    module Queue

      extend ActiveSupport::Concern



      included do

        # If a job produces an error it may block a stage from advancing

        # forever. To prevent this from happening we prevent jobs from going to

        # the dead queue. This does mean some resources may not be imported, but

        # this is better than a project being stuck in the "import" state

        # forever.

        sidekiq_options queue: 'github_importer', dead: false, retry: 10

      end

    end

  end

end