diff --git a/.gitlab/ci/rules.gitlab-ci.yml b/.gitlab/ci/rules.gitlab-ci.yml
index 22ab62b7ca9..1835b6c945c 100644
--- a/.gitlab/ci/rules.gitlab-ci.yml
+++ b/.gitlab/ci/rules.gitlab-ci.yml
@@ -45,10 +45,10 @@
   if: '($CI_MERGE_REQUEST_EVENT_TYPE == "merged_result" || $CI_MERGE_REQUEST_EVENT_TYPE == "detached") && $CI_MERGE_REQUEST_LABELS =~ /pipeline:mr-approved/'
 
 .if-merge-request-approved-and-specific-devops-stage: &if-merge-request-approved-and-specific-devops-stage
-  if: '($CI_MERGE_REQUEST_EVENT_TYPE == "merged_result" || $CI_MERGE_REQUEST_EVENT_TYPE == "detached") && ($CI_MERGE_REQUEST_LABELS =~ /pipeline:mr-approved/ && $CI_MERGE_REQUEST_LABELS =~ /devops::create/)'
+  if: '($CI_MERGE_REQUEST_EVENT_TYPE == "merged_result" || $CI_MERGE_REQUEST_EVENT_TYPE == "detached") && ($CI_MERGE_REQUEST_LABELS =~ /pipeline:mr-approved/ && $CI_MERGE_REQUEST_LABELS =~ /devops::(create|govern|manage)/)'
 
 .if-merge-request-and-specific-devops-stage: &if-merge-request-and-specific-devops-stage
-  if: '($CI_MERGE_REQUEST_EVENT_TYPE == "merged_result" || $CI_MERGE_REQUEST_EVENT_TYPE == "detached") && $CI_MERGE_REQUEST_LABELS =~ /devops::create/'
+  if: '($CI_MERGE_REQUEST_EVENT_TYPE == "merged_result" || $CI_MERGE_REQUEST_EVENT_TYPE == "detached") && $CI_MERGE_REQUEST_LABELS =~ /devops::(create|govern|manage)/'
 
 .if-merge-request-not-approved: &if-merge-request-not-approved
   if: '($CI_MERGE_REQUEST_EVENT_TYPE == "merged_result" || $CI_MERGE_REQUEST_EVENT_TYPE == "detached") && $CI_MERGE_REQUEST_LABELS !~ /pipeline:mr-approved/'
diff --git a/app/models/commit_status.rb b/app/models/commit_status.rb
index 6109ead9865..1695b407daf 100644
--- a/app/models/commit_status.rb
+++ b/app/models/commit_status.rb
@@ -8,6 +8,18 @@ class CommitStatus < Ci::ApplicationRecord
   include Presentable
   include BulkInsertableAssociations
   include TaggableQueries
+  include IgnorableColumns
+
+  ignore_columns %i[
+    auto_canceled_by_id_convert_to_bigint
+    commit_id_convert_to_bigint
+    erased_by_id_convert_to_bigint
+    project_id_convert_to_bigint
+    runner_id_convert_to_bigint
+    trigger_request_id_convert_to_bigint
+    upstream_pipeline_id_convert_to_bigint
+    user_id_convert_to_bigint
+  ], remove_with: '17.0', remove_after: '2024-04-22'
 
   self.table_name = :p_ci_builds
   self.sequence_name = :ci_builds_id_seq
diff --git a/app/models/ml/candidate.rb b/app/models/ml/candidate.rb
index 6f4728a1d98..70eaab8c0ab 100644
--- a/app/models/ml/candidate.rb
+++ b/app/models/ml/candidate.rb
@@ -12,12 +12,14 @@ module Ml
 
     validates :eid, :experiment, presence: true
     validates :status, inclusion: { in: statuses.keys }
+    validates :model_version_id, uniqueness: { allow_nil: true }
 
     belongs_to :experiment, class_name: 'Ml::Experiment'
     belongs_to :user
     belongs_to :package, class_name: 'Packages::Package'
     belongs_to :project
     belongs_to :ci_build, class_name: 'Ci::Build', optional: true
+    belongs_to :model_version, class_name: 'Ml::ModelVersion', optional: true, inverse_of: :candidate
     has_many :metrics, class_name: 'Ml::CandidateMetric'
     has_many :params, class_name: 'Ml::CandidateParam'
     has_many :metadata, class_name: 'Ml::CandidateMetadata'
diff --git a/app/models/ml/model_version.rb b/app/models/ml/model_version.rb
index 36272cb33de..ec55fa9e2ed 100644
--- a/app/models/ml/model_version.rb
+++ b/app/models/ml/model_version.rb
@@ -20,6 +20,7 @@ module Ml
     belongs_to :model, class_name: 'Ml::Model'
     belongs_to :project
     belongs_to :package, class_name: 'Packages::MlModel::Package', optional: true
+    has_one :candidate, class_name: 'Ml::Candidate'
 
     delegate :name, to: :model
 
diff --git a/app/services/ml/create_candidate_service.rb b/app/services/ml/create_candidate_service.rb
new file mode 100644
index 00000000000..53913c3fb19
--- /dev/null
+++ b/app/services/ml/create_candidate_service.rb
@@ -0,0 +1,37 @@
+# frozen_string_literal: true
+
+module Ml
+  class CreateCandidateService
+    def initialize(experiment, params = {})
+      @experiment = experiment
+      @name = params[:name]
+      @user = params[:user]
+      @start_time = params[:start_time]
+      @model_version = params[:model_version]
+    end
+
+    def execute
+      Ml::Candidate.create!(
+        experiment: experiment,
+        project: experiment.project,
+        name: candidate_name,
+        start_time: start_time || 0,
+        user: user,
+        model_version: model_version
+      )
+    end
+
+    private
+
+    def candidate_name
+      name.presence || random_candidate_name
+    end
+
+    def random_candidate_name
+      parts = Array.new(3).map { FFaker::Animal.common_name.downcase.delete(' ') } << rand(10000)
+      parts.join('-').truncate(255)
+    end
+
+    attr_reader :name, :user, :experiment, :start_time, :model_version
+  end
+end
diff --git a/app/services/ml/experiment_tracking/candidate_repository.rb b/app/services/ml/experiment_tracking/candidate_repository.rb
index 436f06e3ca5..8739379912a 100644
--- a/app/services/ml/experiment_tracking/candidate_repository.rb
+++ b/app/services/ml/experiment_tracking/candidate_repository.rb
@@ -15,12 +15,13 @@ module Ml
       end
 
       def create!(experiment, start_time, tags = nil, name = nil)
-        candidate = experiment.candidates.create!(
+        create_params = {
+          start_time: start_time,
           user: user,
-          name: candidate_name(name, tags),
-          project: project,
-          start_time: start_time || 0
-        )
+          name: candidate_name(name, tags)
+        }
+
+        candidate = Ml::CreateCandidateService.new(experiment, create_params).execute
 
         add_tags(candidate, tags)
 
@@ -103,17 +104,12 @@ module Ml
       end
 
       def candidate_name(name, tags)
-        name.presence || candidate_name_from_tags(tags) || random_candidate_name
+        name.presence || candidate_name_from_tags(tags)
       end
 
       def candidate_name_from_tags(tags)
         tags&.detect { |t| t[:key] == 'mlflow.runName' }&.dig(:value)
       end
-
-      def random_candidate_name
-        parts = Array.new(3).map { FFaker::Animal.common_name.downcase.delete(' ') } << rand(10000)
-        parts.join('-').truncate(255)
-      end
     end
   end
 end
diff --git a/app/services/ml/find_or_create_model_version_service.rb b/app/services/ml/find_or_create_model_version_service.rb
index b84c1cec073..cccbe37dbb1 100644
--- a/app/services/ml/find_or_create_model_version_service.rb
+++ b/app/services/ml/find_or_create_model_version_service.rb
@@ -12,7 +12,15 @@ module Ml
 
     def execute
       model = Ml::FindOrCreateModelService.new(project, name).execute
-      Ml::ModelVersion.find_or_create!(model, version, package, description)
+
+      model_version = Ml::ModelVersion.find_or_create!(model, version, package, description)
+
+      model_version.candidate = ::Ml::CreateCandidateService.new(
+        model.default_experiment,
+        { model_version: model_version }
+      ).execute
+
+      model_version
     end
 
     private
diff --git a/db/docs/approval_group_rules.yml b/db/docs/approval_group_rules.yml
index 35f75338480..b9dab08c5df 100644
--- a/db/docs/approval_group_rules.yml
+++ b/db/docs/approval_group_rules.yml
@@ -1,9 +1,9 @@
 ---
 table_name: approval_group_rules
 classes:
-  - ApprovalRules::ApprovalGroupRule
+- ApprovalRules::ApprovalGroupRule
 feature_categories:
-  - source_code_management
+- source_code_management
 description: Keeps approval group rules
 introduced_by_url: https://gitlab.com/gitlab-org/gitlab/-/merge_requests/132651
 milestone: '16.5'
diff --git a/db/migrate/20231020112541_add_column_model_version_id_to_ml_candidates.rb b/db/migrate/20231020112541_add_column_model_version_id_to_ml_candidates.rb
new file mode 100644
index 00000000000..7bfe78c4ebd
--- /dev/null
+++ b/db/migrate/20231020112541_add_column_model_version_id_to_ml_candidates.rb
@@ -0,0 +1,7 @@
+# frozen_string_literal: true
+
+class AddColumnModelVersionIdToMlCandidates < Gitlab::Database::Migration[2.1]
+  def change
+    add_column :ml_candidates, :model_version_id, :bigint, null: true
+  end
+end
diff --git a/db/migrate/20231023114006_add_index_on_model_version_id_to_ml_candidates.rb b/db/migrate/20231023114006_add_index_on_model_version_id_to_ml_candidates.rb
new file mode 100644
index 00000000000..598600b8539
--- /dev/null
+++ b/db/migrate/20231023114006_add_index_on_model_version_id_to_ml_candidates.rb
@@ -0,0 +1,15 @@
+# frozen_string_literal: true
+
+class AddIndexOnModelVersionIdToMlCandidates < Gitlab::Database::Migration[2.1]
+  disable_ddl_transaction!
+
+  INDEX_NAME = 'index_ml_candidates_on_model_version_id'
+
+  def up
+    add_concurrent_index :ml_candidates, :model_version_id, name: INDEX_NAME, unique: true
+  end
+
+  def down
+    remove_concurrent_index_by_name :ml_candidates, name: INDEX_NAME
+  end
+end
diff --git a/db/migrate/20231023114551_add_fk_on_ml_candidates_to_ml_model_versions.rb b/db/migrate/20231023114551_add_fk_on_ml_candidates_to_ml_model_versions.rb
new file mode 100644
index 00000000000..0d625a54656
--- /dev/null
+++ b/db/migrate/20231023114551_add_fk_on_ml_candidates_to_ml_model_versions.rb
@@ -0,0 +1,15 @@
+# frozen_string_literal: true
+
+class AddFkOnMlCandidatesToMlModelVersions < Gitlab::Database::Migration[2.1]
+  disable_ddl_transaction!
+
+  def up
+    add_concurrent_foreign_key(:ml_candidates, :ml_model_versions, column: :model_version_id, on_delete: :cascade)
+  end
+
+  def down
+    with_lock_retries do
+      remove_foreign_key_if_exists(:ml_model_versions, column: :model_version_id, on_delete: :cascade)
+    end
+  end
+end
diff --git a/db/post_migrate/20231023083349_init_conversion_for_p_ci_builds.rb b/db/post_migrate/20231023083349_init_conversion_for_p_ci_builds.rb
new file mode 100644
index 00000000000..886e8c85599
--- /dev/null
+++ b/db/post_migrate/20231023083349_init_conversion_for_p_ci_builds.rb
@@ -0,0 +1,37 @@
+# frozen_string_literal: true
+
+class InitConversionForPCiBuilds < Gitlab::Database::Migration[2.1]
+  include ::Gitlab::Database::MigrationHelpers::WraparoundAutovacuum
+
+  disable_ddl_transaction!
+
+  TABLE_NAME = :p_ci_builds
+  COLUMN_NAMES = %i[
+    auto_canceled_by_id
+    commit_id
+    erased_by_id
+    project_id
+    runner_id
+    trigger_request_id
+    upstream_pipeline_id
+    user_id
+  ]
+
+  def up
+    return if should_skip?
+
+    initialize_conversion_of_integer_to_bigint(TABLE_NAME, COLUMN_NAMES)
+  end
+
+  def down
+    return if should_skip?
+
+    revert_initialize_conversion_of_integer_to_bigint(TABLE_NAME, COLUMN_NAMES)
+  end
+
+  private
+
+  def should_skip?
+    !can_execute_on?(TABLE_NAME)
+  end
+end
diff --git a/db/schema_migrations/20231020112541 b/db/schema_migrations/20231020112541
new file mode 100644
index 00000000000..f385bb06bd6
--- /dev/null
+++ b/db/schema_migrations/20231020112541
@@ -0,0 +1 @@
+16a0b32619e6b28c49fc2e9e609970ac61582c295c2ed281c531e407de2af216
\ No newline at end of file
diff --git a/db/schema_migrations/20231023083349 b/db/schema_migrations/20231023083349
new file mode 100644
index 00000000000..d3d7e4e45fc
--- /dev/null
+++ b/db/schema_migrations/20231023083349
@@ -0,0 +1 @@
+1b0cd52ccf99a477f39168cdb6b719d5b64f6110a7fd9df0a6f200c6ff9c0237
\ No newline at end of file
diff --git a/db/schema_migrations/20231023114006 b/db/schema_migrations/20231023114006
new file mode 100644
index 00000000000..bc17ae9b852
--- /dev/null
+++ b/db/schema_migrations/20231023114006
@@ -0,0 +1 @@
+030809f5519906dbdcdf3b8fd35a7181ca2c9ec1bdca745997aa14f24ee6ac6d
\ No newline at end of file
diff --git a/db/schema_migrations/20231023114551 b/db/schema_migrations/20231023114551
new file mode 100644
index 00000000000..a53b51b53bd
--- /dev/null
+++ b/db/schema_migrations/20231023114551
@@ -0,0 +1 @@
+df2937c8e70fde85677ef150ed6c2445d06b8a2f7113f58a8908a81ece449d75
\ No newline at end of file
diff --git a/db/structure.sql b/db/structure.sql
index 60387cc945f..177e2d026fb 100644
--- a/db/structure.sql
+++ b/db/structure.sql
@@ -298,6 +298,22 @@ BEGIN
 END;
 $$;
 
+CREATE FUNCTION trigger_10ee1357e825() RETURNS trigger
+    LANGUAGE plpgsql
+    AS $$
+BEGIN
+  NEW."auto_canceled_by_id_convert_to_bigint" := NEW."auto_canceled_by_id";
+  NEW."commit_id_convert_to_bigint" := NEW."commit_id";
+  NEW."erased_by_id_convert_to_bigint" := NEW."erased_by_id";
+  NEW."project_id_convert_to_bigint" := NEW."project_id";
+  NEW."runner_id_convert_to_bigint" := NEW."runner_id";
+  NEW."trigger_request_id_convert_to_bigint" := NEW."trigger_request_id";
+  NEW."upstream_pipeline_id_convert_to_bigint" := NEW."upstream_pipeline_id";
+  NEW."user_id_convert_to_bigint" := NEW."user_id";
+  RETURN NEW;
+END;
+$$;
+
 CREATE FUNCTION trigger_1bd97da9c1a4() RETURNS trigger
     LANGUAGE plpgsql
     AS $$
@@ -13379,6 +13395,14 @@ CREATE TABLE p_ci_builds (
     stage_id bigint,
     partition_id bigint NOT NULL,
     auto_canceled_by_partition_id bigint DEFAULT 100 NOT NULL,
+    auto_canceled_by_id_convert_to_bigint bigint,
+    commit_id_convert_to_bigint bigint,
+    erased_by_id_convert_to_bigint bigint,
+    project_id_convert_to_bigint bigint,
+    runner_id_convert_to_bigint bigint,
+    trigger_request_id_convert_to_bigint bigint,
+    upstream_pipeline_id_convert_to_bigint bigint,
+    user_id_convert_to_bigint bigint,
     CONSTRAINT check_1e2fbd1b39 CHECK ((lock_version IS NOT NULL))
 )
 PARTITION BY LIST (partition_id);
@@ -13429,6 +13453,14 @@ CREATE TABLE ci_builds (
     stage_id bigint,
     partition_id bigint NOT NULL,
     auto_canceled_by_partition_id bigint DEFAULT 100 NOT NULL,
+    auto_canceled_by_id_convert_to_bigint bigint,
+    commit_id_convert_to_bigint bigint,
+    erased_by_id_convert_to_bigint bigint,
+    project_id_convert_to_bigint bigint,
+    runner_id_convert_to_bigint bigint,
+    trigger_request_id_convert_to_bigint bigint,
+    upstream_pipeline_id_convert_to_bigint bigint,
+    user_id_convert_to_bigint bigint,
     CONSTRAINT check_1e2fbd1b39 CHECK ((lock_version IS NOT NULL))
 );
 
@@ -18883,6 +18915,7 @@ CREATE TABLE ml_candidates (
     project_id bigint,
     internal_id bigint,
     ci_build_id bigint,
+    model_version_id bigint,
     CONSTRAINT check_25e6c65051 CHECK ((char_length(name) <= 255)),
     CONSTRAINT check_cd160587d4 CHECK ((eid IS NOT NULL))
 );
@@ -33229,6 +33262,8 @@ CREATE INDEX index_ml_candidates_on_ci_build_id ON ml_candidates USING btree (ci
 
 CREATE UNIQUE INDEX index_ml_candidates_on_experiment_id_and_eid ON ml_candidates USING btree (experiment_id, eid);
 
+CREATE UNIQUE INDEX index_ml_candidates_on_model_version_id ON ml_candidates USING btree (model_version_id);
+
 CREATE INDEX index_ml_candidates_on_package_id ON ml_candidates USING btree (package_id);
 
 CREATE INDEX index_ml_candidates_on_project_id ON ml_candidates USING btree (project_id);
@@ -36747,6 +36782,8 @@ CREATE TRIGGER tags_loose_fk_trigger AFTER DELETE ON tags REFERENCING OLD TABLE
 
 CREATE TRIGGER trigger_07bc3c48f407 BEFORE INSERT OR UPDATE ON ci_stages FOR EACH ROW EXECUTE FUNCTION trigger_07bc3c48f407();
 
+CREATE TRIGGER trigger_10ee1357e825 BEFORE INSERT OR UPDATE ON p_ci_builds FOR EACH ROW EXECUTE FUNCTION trigger_10ee1357e825();
+
 CREATE TRIGGER trigger_1bd97da9c1a4 BEFORE INSERT OR UPDATE ON ci_pipelines FOR EACH ROW EXECUTE FUNCTION trigger_1bd97da9c1a4();
 
 CREATE TRIGGER trigger_68d7b6653c7d BEFORE INSERT OR UPDATE ON ci_sources_pipelines FOR EACH ROW EXECUTE FUNCTION trigger_68d7b6653c7d();
@@ -37828,6 +37865,9 @@ ALTER TABLE ONLY namespaces
 ALTER TABLE ONLY fork_networks
     ADD CONSTRAINT fk_e7b436b2b5 FOREIGN KEY (root_project_id) REFERENCES projects(id) ON DELETE SET NULL;
 
+ALTER TABLE ONLY ml_candidates
+    ADD CONSTRAINT fk_e86e0bfa5a FOREIGN KEY (model_version_id) REFERENCES ml_model_versions(id) ON DELETE CASCADE;
+
 ALTER TABLE ONLY integrations
     ADD CONSTRAINT fk_e8fe908a34 FOREIGN KEY (group_id) REFERENCES namespaces(id) ON DELETE CASCADE;
 
diff --git a/doc/administration/gitaly/configure_gitaly.md b/doc/administration/gitaly/configure_gitaly.md
index f62f0a5a4e2..c4f064b5eba 100644
--- a/doc/administration/gitaly/configure_gitaly.md
+++ b/doc/administration/gitaly/configure_gitaly.md
@@ -361,7 +361,7 @@ Configure Gitaly server in one of two ways:
 WARNING:
 If directly copying repository data from a GitLab server to Gitaly, ensure that the metadata file,
 default path `/var/opt/gitlab/git-data/repositories/.gitaly-metadata`, is not included in the transfer.
-Copying this file causes GitLab to use the [Rugged patches](index.md#direct-access-to-git-in-gitlab) for repositories hosted on the Gitaly server,
+Copying this file causes GitLab to use the direct disk access to repositories hosted on the Gitaly server,
 leading to `Error creating pipeline` and `Commit not found` errors, or stale data.
 
 ### Configure Gitaly clients
diff --git a/doc/administration/gitaly/index.md b/doc/administration/gitaly/index.md
index 46f6a5829c8..6784ff4d970 100644
--- a/doc/administration/gitaly/index.md
+++ b/doc/administration/gitaly/index.md
@@ -587,92 +587,6 @@ off Gitaly Cluster to a sharded Gitaly instance:
 1. [Move the repositories](../operations/moving_repositories.md#moving-repositories) to the newly created storage. You can
    move them by shard or by group, which gives you the opportunity to spread them over multiple Gitaly servers.
 
-## Direct access to Git in GitLab
-
-Direct access to Git uses code in GitLab known as the "Rugged patches".
-
-Before Gitaly existed, what are now Gitaly clients accessed Git repositories directly, either:
-
-- On a local disk in the case of a single-machine Linux package installation.
-- Using NFS in the case of a horizontally-scaled GitLab installation.
-
-In addition to running plain `git` commands, GitLab used a Ruby library called
-[Rugged](https://github.com/libgit2/rugged). Rugged is a wrapper around
-[libgit2](https://libgit2.org/), a stand-alone implementation of Git in the form of a C library.
-
-Over time it became clear that Rugged, particularly in combination with
-[Unicorn](https://yhbt.net/unicorn/), is extremely efficient. Because `libgit2` is a library and
-not an external process, there was very little overhead between:
-
-- GitLab application code that tried to look up data in Git repositories.
-- The Git implementation itself.
-
-Because the combination of Rugged and Unicorn was so efficient, the GitLab application code ended up
-with lots of duplicate Git object lookups. For example, looking up the default branch commit a dozen
-times in one request. We could write inefficient code without poor performance.
-
-When we migrated these Git lookups to Gitaly calls, we suddenly had a much higher fixed cost per Git
-lookup. Even when Gitaly is able to re-use an already-running `git` process (for example, to look up
-a commit), you still have:
-
-- The cost of a network roundtrip to Gitaly.
-- Inside Gitaly, a write/read roundtrip on the Unix pipes that connect Gitaly to the `git` process.
-
-Using GitLab.com to measure, we reduced the number of Gitaly calls per request until we no longer felt
-the efficiency loss of losing Rugged. It also helped that we run Gitaly itself directly on the Git
-file servers, rather than by using NFS mounts. This gave us a speed boost that counteracted the
-negative effect of not using Rugged anymore.
-
-Unfortunately, other deployments of GitLab could not remove NFS like we did on GitLab.com, and they
-got the worst of both worlds:
-
-- The slowness of NFS.
-- The increased inherent overhead of Gitaly.
-
-The code removed from GitLab during the Gitaly migration project affected these deployments. As a
-performance workaround for these NFS-based deployments, we re-introduced some of the old Rugged
-code. This re-introduced code is informally referred to as the "Rugged patches".
-
-### Automatic detection
-
-> Automatic detection for Rugged [disabled](https://gitlab.com/gitlab-org/gitlab/-/merge_requests/95445) in GitLab 15.3.
-
-FLAG:
-On self-managed GitLab, by default automatic detection of whether Rugged should be used (per storage) is not available.
-To make it available, an administrator can [disable the feature flag](../../administration/feature_flags.md) named
-`skip_rugged_auto_detect`.
-
-The Ruby methods that perform direct Git access are behind
-[feature flags](../../development/gitaly.md#legacy-rugged-code), disabled by default. It wasn't
-convenient to set feature flags to get the best performance, so we added an automatic mechanism that
-enables direct Git access.
-
-When GitLab calls a function that has a "Rugged patch", it performs two checks:
-
-- Is the feature flag for this patch set in the database? If so, the feature flag setting controls
-  the GitLab use of "Rugged patch" code.
-- If the feature flag is not set, GitLab tries accessing the file system underneath the
-  Gitaly server directly. If it can, it uses the "Rugged patch":
-  - If using Puma and [thread count](../../install/requirements.md#puma-threads) is set
-    to `1`.
-
-The result of these checks is cached.
-
-To see if GitLab can access the repository file system directly, we use the following heuristic:
-
-- Gitaly ensures that the file system has a metadata file in its root with a UUID in it.
-- Gitaly reports this UUID to GitLab by using the `ServerInfo` RPC.
-- GitLab Rails tries to read the metadata file directly. If it exists, and if the UUIDs match,
-  assume we have direct access.
-
-Direct Git access is:
-
-- [Disabled](https://gitlab.com/gitlab-org/gitlab/-/merge_requests/95445) by default in GitLab 15.3 and later for
-  compatibility with [Praefect-generated replica paths](#praefect-generated-replica-paths-gitlab-150-and-later). It
-  can be enabled if Rugged [feature flags](../../development/gitaly.md#legacy-rugged-code) are enabled.
-- Enabled by default in GitLab 15.2 and earlier because it fills in the correct repository paths in the GitLab
-  configuration file `config/gitlab.yml`. This satisfies the UUID check.
-
 ### Transition to Gitaly Cluster
 
 For the sake of removing complexity, we must remove direct Git access in GitLab. However, we can't
diff --git a/doc/administration/monitoring/prometheus/gitlab_metrics.md b/doc/administration/monitoring/prometheus/gitlab_metrics.md
index 16287741e6a..f6ee2961ce2 100644
--- a/doc/administration/monitoring/prometheus/gitlab_metrics.md
+++ b/doc/administration/monitoring/prometheus/gitlab_metrics.md
@@ -243,7 +243,6 @@ configuration option in `gitlab.yml`. These metrics are served from the
 | `geo_cursor_last_event_timestamp`              | Gauge   | 10.2  | Last UNIX timestamp of the event log processed by the secondary | `url` |
 | `geo_status_failed_total`                      | Counter | 10.2  | Number of times retrieving the status from the Geo Node failed | `url` |
 | `geo_last_successful_status_check_timestamp`   | Gauge   | 10.2  | Last timestamp when the status was successfully updated | `url` |
-| `geo_job_artifacts_synced_missing_on_primary`  | Gauge   | 10.7  | Number of job artifacts marked as synced due to the file missing on the primary | `url` |
 | `geo_package_files`                            | Gauge   | 13.0  | Number of package files on primary | `url` |
 | `geo_package_files_checksummed`                | Gauge   | 13.0  | Number of package files checksummed on primary | `url` |
 | `geo_package_files_checksum_failed`            | Gauge   | 13.0  | Number of package files failed to calculate the checksum on primary | `url` |
diff --git a/doc/administration/operations/puma.md b/doc/administration/operations/puma.md
index f16f1ac46ae..89f1574697f 100644
--- a/doc/administration/operations/puma.md
+++ b/doc/administration/operations/puma.md
@@ -140,37 +140,6 @@ When running Puma in single mode, some features are not supported:
 
 For more information, see [epic 5303](https://gitlab.com/groups/gitlab-org/-/epics/5303).
 
-## Performance caveat when using Puma with Rugged
-
-For deployments where NFS is used to store Git repositories, GitLab uses
-[direct Git access](../gitaly/index.md#direct-access-to-git-in-gitlab) to improve performance by using
-[Rugged](https://github.com/libgit2/rugged).
-
-Rugged usage is automatically enabled if direct Git access [is available](../gitaly/index.md#automatic-detection) and
-Puma is running single threaded, unless it is disabled by a [feature flag](../../development/gitaly.md#legacy-rugged-code).
-
-MRI Ruby uses a Global VM Lock (GVL). GVL allows MRI Ruby to be multi-threaded, but running at
-most on a single core.
-
-Git includes intensive I/O operations. When Rugged uses a thread for a long period of time,
-other threads that might be processing requests can starve. Puma running in single thread mode
-does not have this issue, because concurrently at most one request is being processed.
-
-GitLab is working to remove Rugged usage. Even though performance without Rugged
-is acceptable today, in some cases it might be still beneficial to run with it.
-
-Given the caveat of running Rugged with multi-threaded Puma, and acceptable
-performance of Gitaly, we disable Rugged usage if Puma multi-threaded is
-used (when Puma is configured to run with more than one thread).
-
-This default behavior may not be the optimal configuration in some situations. If Rugged
-plays an important role in your deployment, we suggest you benchmark to find the
-optimal configuration:
-
-- The safest option is to start with single-threaded Puma.
-- To force Rugged to be used with multi-threaded Puma, you can use a
-  [feature flag](../../development/gitaly.md#legacy-rugged-code).
-
 ## Configuring Puma to listen over SSL
 
 Puma, when deployed with a Linux package installation, listens over a Unix socket by
diff --git a/doc/administration/sidekiq/index.md b/doc/administration/sidekiq/index.md
index a27723faa4a..0a7974c9622 100644
--- a/doc/administration/sidekiq/index.md
+++ b/doc/administration/sidekiq/index.md
@@ -356,20 +356,6 @@ To enable LDAP with the synchronization worker for Sidekiq:
 
 If you use [SAML Group Sync](../../user/group/saml_sso/group_sync.md), you must configure [SAML Groups](../../integration/saml.md#configure-users-based-on-saml-group-membership) on all your Sidekiq nodes.
 
-## Disable Rugged
-
-Calls into Rugged, Ruby bindings for `libgit2`, [lock the Sidekiq processes (GVL)](https://silverhammermba.github.io/emberb/c/#c-in-ruby-threads),
-blocking all jobs on that worker from proceeding. If Rugged calls performed by Sidekiq are slow, this can cause significant delays in
-background task processing.
-
-By default, Rugged is used when Git repository data is stored on local storage or on an NFS mount.
-Using Rugged is recommended when using NFS, but if
-you are using local storage, disabling Rugged can improve Sidekiq performance:
-
-```shell
-sudo gitlab-rake gitlab:features:disable_rugged
-```
-
 ## Related topics
 
 - [Extra Sidekiq processes](extra_sidekiq_processes.md)
diff --git a/doc/api/geo_nodes.md b/doc/api/geo_nodes.md
index 3f7fd537abf..c376d7a6774 100644
--- a/doc/api/geo_nodes.md
+++ b/doc/api/geo_nodes.md
@@ -332,7 +332,6 @@ Example response:
     "job_artifacts_count": 2,
     "job_artifacts_synced_count": null,
     "job_artifacts_failed_count": null,
-    "job_artifacts_synced_missing_on_primary_count": 0,
     "job_artifacts_synced_in_percentage": "0.00%",
     "projects_count": 41,
     "repositories_count": 41,
@@ -470,7 +469,6 @@ Example response:
     "job_artifacts_verification_failed_count": 0,
     "job_artifacts_synced_in_percentage": "100.00%",
     "job_artifacts_verified_in_percentage": "100.00%",
-    "job_artifacts_synced_missing_on_primary_count": 0,
     "ci_secure_files_count": 5,
     "ci_secure_files_checksum_total_count": 5,
     "ci_secure_files_checksummed_count": 5,
@@ -483,7 +481,6 @@ Example response:
     "ci_secure_files_verification_failed_count": 0,
     "ci_secure_files_synced_in_percentage": "100.00%",
     "ci_secure_files_verified_in_percentage": "100.00%",
-    "ci_secure_files_synced_missing_on_primary_count": 0,
     "dependency_proxy_blobs_count": 5,
     "dependency_proxy_blobs_checksum_total_count": 5,
     "dependency_proxy_blobs_checksummed_count": 5,
@@ -496,13 +493,11 @@ Example response:
     "dependency_proxy_blobs_verification_failed_count": 0,
     "dependency_proxy_blobs_synced_in_percentage": "100.00%",
     "dependency_proxy_blobs_verified_in_percentage": "100.00%",
-    "dependency_proxy_blobs_synced_missing_on_primary_count": 0,
     "container_repositories_count": 5,
     "container_repositories_synced_count": 5,
     "container_repositories_failed_count": 0,
     "container_repositories_registry_count": 5,
     "container_repositories_synced_in_percentage": "100.00%",
-    "container_repositories_synced_missing_on_primary_count": 0,
     "container_repositories_checksum_total_count": 0,
     "container_repositories_checksummed_count": 0,
     "container_repositories_checksum_failed_count": 0,
@@ -569,7 +564,6 @@ Example response:
     "job_artifacts_count": 2,
     "job_artifacts_synced_count": 1,
     "job_artifacts_failed_count": 1,
-    "job_artifacts_synced_missing_on_primary_count": 0,
     "job_artifacts_synced_in_percentage": "50.00%",
     "design_management_repositories_count": 5,
     "design_management_repositories_synced_count": 5,
@@ -695,7 +689,6 @@ Example response:
     "job_artifacts_verification_failed_count": 0,
     "job_artifacts_synced_in_percentage": "100.00%",
     "job_artifacts_verified_in_percentage": "100.00%",
-    "job_artifacts_synced_missing_on_primary_count": 0,
     "dependency_proxy_blobs_count": 5,
     "dependency_proxy_blobs_checksum_total_count": 5,
     "dependency_proxy_blobs_checksummed_count": 5,
@@ -708,13 +701,11 @@ Example response:
     "dependency_proxy_blobs_verification_failed_count": 0,
     "dependency_proxy_blobs_synced_in_percentage": "100.00%",
     "dependency_proxy_blobs_verified_in_percentage": "100.00%",
-    "dependency_proxy_blobs_synced_missing_on_primary_count": 0,
     "container_repositories_count": 5,
     "container_repositories_synced_count": 5,
     "container_repositories_failed_count": 0,
     "container_repositories_registry_count": 5,
     "container_repositories_synced_in_percentage": "100.00%",
-    "container_repositories_synced_missing_on_primary_count": 0,
     "container_repositories_checksum_total_count": 0,
     "container_repositories_checksummed_count": 0,
     "container_repositories_checksum_failed_count": 0,
@@ -785,7 +776,6 @@ Example response:
   "job_artifacts_count": 2,
   "job_artifacts_synced_count": 1,
   "job_artifacts_failed_count": 1,
-  "job_artifacts_synced_missing_on_primary_count": 0,
   "job_artifacts_synced_in_percentage": "50.00%",
   "projects_count": 41,
   "repositories_count": 41,
@@ -896,7 +886,6 @@ Example response:
   "job_artifacts_verification_failed_count": 0,
   "job_artifacts_synced_in_percentage": "100.00%",
   "job_artifacts_verified_in_percentage": "100.00%",
-  "job_artifacts_synced_missing_on_primary_count": 0,
   "ci_secure_files_count": 5,
   "ci_secure_files_checksum_total_count": 5,
   "ci_secure_files_checksummed_count": 5,
@@ -909,7 +898,6 @@ Example response:
   "ci_secure_files_verification_failed_count": 0,
   "ci_secure_files_synced_in_percentage": "100.00%",
   "ci_secure_files_verified_in_percentage": "100.00%",
-  "ci_secure_files_synced_missing_on_primary_count": 0,
   "dependency_proxy_blobs_count": 5,
   "dependency_proxy_blobs_checksum_total_count": 5,
   "dependency_proxy_blobs_checksummed_count": 5,
@@ -922,13 +910,11 @@ Example response:
   "dependency_proxy_blobs_verification_failed_count": 0,
   "dependency_proxy_blobs_synced_in_percentage": "100.00%",
   "dependency_proxy_blobs_verified_in_percentage": "100.00%",
-  "dependency_proxy_blobs_synced_missing_on_primary_count": 0,
   "container_repositories_count": 5,
   "container_repositories_synced_count": 5,
   "container_repositories_failed_count": 0,
   "container_repositories_registry_count": 5,
   "container_repositories_synced_in_percentage": "100.00%",
-  "container_repositories_synced_missing_on_primary_count": 0,
   "container_repositories_checksum_total_count": 0,
   "container_repositories_checksummed_count": 0,
   "container_repositories_checksum_failed_count": 0,
diff --git a/doc/api/geo_sites.md b/doc/api/geo_sites.md
index eaf813ae201..95691960a78 100644
--- a/doc/api/geo_sites.md
+++ b/doc/api/geo_sites.md
@@ -292,7 +292,6 @@ Example response:
 [
   {
     "geo_node_id": 1,
-    "job_artifacts_synced_missing_on_primary_count": null,
     "projects_count": 19,
     "container_repositories_replication_enabled": null,
     "lfs_objects_count": 0,
@@ -510,7 +509,6 @@ Example response:
   },
   {
     "geo_node_id": 2,
-    "job_artifacts_synced_missing_on_primary_count": null,
     "projects_count": 19,
     "container_repositories_replication_enabled": null,
     "lfs_objects_count": 0,
@@ -744,7 +742,6 @@ Example response:
 ```json
   {
   "geo_node_id": 2,
-  "job_artifacts_synced_missing_on_primary_count": null,
   "projects_count": 19,
   "container_repositories_replication_enabled": null,
   "lfs_objects_count": 0,
diff --git a/doc/development/backend/create_source_code_be/gitaly_touch_points.md b/doc/development/backend/create_source_code_be/gitaly_touch_points.md
index c689af2f150..98607c7f6c7 100644
--- a/doc/development/backend/create_source_code_be/gitaly_touch_points.md
+++ b/doc/development/backend/create_source_code_be/gitaly_touch_points.md
@@ -19,9 +19,3 @@ All access to Gitaly from other parts of GitLab are through Create: Source Code
 After a call is made to Gitaly, Git `commit` information is stored in memory. This information is wrapped by the [Ruby `Commit` Model](https://gitlab.com/gitlab-org/gitlab/-/blob/master/app/models/commit.rb), which is a wrapper around [`Gitlab::Git::Commit`](https://gitlab.com/gitlab-org/gitlab/-/blob/master/lib/gitlab/git/commit.rb).
 
 The `Commit` model acts like an ActiveRecord object, but it does not have a PostgreSQL backend. Instead, it maps back to Gitaly RPCs.
-
-## Rugged Patches
-
-Historically in GitLab, access to the server-based `git` repositories was provided through the [rugged](https://github.com/libgit2/rugged) RubyGem, which provides Ruby bindings to `libgit2`. This was further extended by what is termed "Rugged Patches", [a set of extensions to the Rugged library](https://gitlab.com/gitlab-org/gitlab-foss/-/issues/57317). Rugged implementations of some of the most commonly-used RPCs can be [enabled via feature flags](../../gitaly.md#legacy-rugged-code).
-
-Rugged access requires the use of a NFS file system, a direction GitLab is moving away from in favor of Gitaly Cluster. Rugged has been proposed for [deprecation and removal](https://gitlab.com/gitlab-org/gitaly/-/issues/1690). Several large customers are still using NFS, and a specific removal date is not planned at this point.
diff --git a/doc/development/gitaly.md b/doc/development/gitaly.md
index e6a853c107e..d23e00748cd 100644
--- a/doc/development/gitaly.md
+++ b/doc/development/gitaly.md
@@ -64,45 +64,6 @@ rm -rf tmp/tests/gitaly
 
 During RSpec tests, the Gitaly instance writes logs to `gitlab/log/gitaly-test.log`.
 
-## Legacy Rugged code
-
-While Gitaly can handle all Git access, many of GitLab customers still
-run Gitaly atop NFS. The legacy Rugged implementation for Git calls may
-be faster than the Gitaly RPC due to N+1 Gitaly calls and other
-reasons. See [the issue](https://gitlab.com/gitlab-org/gitlab-foss/-/issues/57317) for more
-details.
-
-Until GitLab has eliminated most of these inefficiencies or the use of
-NFS is discontinued for Git data, Rugged implementations of some of the
-most commonly-used RPCs can be enabled via feature flags:
-
-- `rugged_find_commit`
-- `rugged_get_tree_entries`
-- `rugged_tree_entry`
-- `rugged_commit_is_ancestor`
-- `rugged_commit_tree_entry`
-- `rugged_list_commits_by_oid`
-
-A convenience Rake task can be used to enable or disable these flags
-all together. To enable:
-
-```shell
-bundle exec rake gitlab:features:enable_rugged
-```
-
-To disable:
-
-```shell
-bundle exec rake gitlab:features:disable_rugged
-```
-
-Most of this code exists in the `lib/gitlab/git/rugged_impl` directory.
-
-NOTE:
-You should *not* have to add or modify code related to Rugged unless explicitly discussed with the
-[Gitaly Team](https://gitlab.com/groups/gl-gitaly/group_members). This code does not work on GitLab.com or other GitLab
-instances that do not use NFS.
-
 ## `TooManyInvocationsError` errors
 
 During development and testing, you may experience `Gitlab::GitalyClient::TooManyInvocationsError` failures.
diff --git a/doc/development/wikis.md b/doc/development/wikis.md
index a814fa76ec9..eca43f6df03 100644
--- a/doc/development/wikis.md
+++ b/doc/development/wikis.md
@@ -28,9 +28,6 @@ Some notable gems that are used for wikis are:
 | Component     | Description                                    | Gem name                       | GitLab project                                                                                          | Upstream project                                                    |
 |:--------------|:-----------------------------------------------|:-------------------------------|:--------------------------------------------------------------------------------------------------------|:--------------------------------------------------------------------|
 | `gitlab`      | Markup renderer, depends on various other gems | `gitlab-markup`                | [`gitlab-org/gitlab-markup`](https://gitlab.com/gitlab-org/gitlab-markup)                               | [`github/markup`](https://github.com/github/markup)                 |
-| `gollum-lib`  | Main Gollum library                            | `gitlab-gollum-lib`            | [`gitlab-org/gollum-lib`](https://gitlab.com/gitlab-org/gollum-lib)                                     | [`gollum/gollum-lib`](https://github.com/gollum/gollum-lib)         |
-|               | Gollum Git adapter for Rugged                  | `gitlab-gollum-rugged_adapter` | [`gitlab-org/gitlab-gollum-rugged_adapter`](https://gitlab.com/gitlab-org/gitlab-gollum-rugged_adapter) | [`gollum/rugged_adapter`](https://github.com/gollum/rugged_adapter) |
-|               | Rugged (also used in Gitaly itself)            | `rugged`                       | -                                                                                                       | [`libgit2/rugged`](https://github.com/libgit2/rugged)               |
 
 ### Notes on Gollum
 
diff --git a/doc/install/requirements.md b/doc/install/requirements.md
index 81244594a59..b02b55757dd 100644
--- a/doc/install/requirements.md
+++ b/doc/install/requirements.md
@@ -227,8 +227,7 @@ optimal settings for your infrastructure.
 
 ### Puma threads
 
-The recommended number of threads is dependent on several factors, including total memory, and use
-of [legacy Rugged code](../administration/gitaly/index.md#direct-access-to-git-in-gitlab).
+The recommended number of threads is dependent on several factors, including total memory.
 
 - If the operating system has a maximum 2 GB of memory, the recommended number of threads is `1`.
   A higher value results in excess swapping, and decrease performance.
diff --git a/doc/user/application_security/container_scanning/index.md b/doc/user/application_security/container_scanning/index.md
index 204c071e728..bfe3dcd9def 100644
--- a/doc/user/application_security/container_scanning/index.md
+++ b/doc/user/application_security/container_scanning/index.md
@@ -276,6 +276,7 @@ including a large number of false positives.
 | `CS_REGISTRY_USER`                  | `$CI_REGISTRY_USER` | Username for accessing a Docker registry requiring authentication. The default is only set if `$CS_IMAGE` resides at [`$CI_REGISTRY`](../../../ci/variables/predefined_variables.md). Not supported when [FIPS mode](../../../development/fips_compliance.md#enable-fips-mode) is enabled. | All |
 | `CS_DOCKERFILE_PATH`              | `Dockerfile`  | The path to the `Dockerfile` to use for generating remediations. By default, the scanner looks for a file named `Dockerfile` in the root directory of the project. You should configure this variable only if your `Dockerfile` is in a non-standard location, such as a subdirectory. See [Solutions for vulnerabilities](#solutions-for-vulnerabilities-auto-remediation) for more details. | All |
 | `CS_QUIET`                     | `""`          | If set, this variable disables output of the [vulnerabilities table](#container-scanning-job-log-format) in the job log. [Introduced](https://gitlab.com/gitlab-org/security-products/analyzers/container-scanning/-/merge_requests/50) in GitLab 15.1. | All |
+| `CS_TRIVY_JAVA_DB`             | `"ghcr.io/aquasecurity/trivy-java-db"` | Specify an alternate location for the [trivy-java-db](https://github.com/aquasecurity/trivy-java-db) vulnerability database. | Trivy |
 | `SECURE_LOG_LEVEL`             | `info`        | Set the minimum logging level. Messages of this logging level or higher are output. From highest to lowest severity, the logging levels are: `fatal`, `error`, `warn`, `info`, `debug`. [Introduced](https://gitlab.com/gitlab-org/gitlab/-/issues/10880) in GitLab 13.1. | All |
 
 ### Supported distributions
@@ -658,6 +659,32 @@ Also:
 
 Scanning images in external private registries is not supported when [FIPS mode](../../../development/fips_compliance.md#enable-fips-mode) is enabled.
 
+#### Create and use a Trivy Java database mirror
+
+When the `trivy` scanner is used and a `jar` file is encountered in a container image being scanned, `trivy` downloads an additional `trivy-java-db` vulnerability database. By default, the `trivy-java-db` database is hosted as an [OCI artifact](https://oras.land/docs/quickstart) at `ghcr.io/aquasecurity/trivy-java-db:1`. If this registry is not accessible, for example in a network-isolated offline GitLab instance, one solution is to mirror the `trivy-java-db` to a container registry that can be accessed in the offline instance:
+
+```yaml
+mirror trivy java db:
+  image:
+    name: ghcr.io/oras-project/oras:v1.1.0
+    entrypoint: [""]
+  script:
+    - oras login -u $CI_REGISTRY_USER -p $CI_REGISTRY_PASSWORD $CI_REGISTRY
+    - oras pull ghcr.io/aquasecurity/trivy-java-db:1
+    - oras push $CI_REGISTRY_IMAGE:1 --config /dev/null:application/vnd.aquasec.trivy.config.v1+json javadb.tar.gz:application/vnd.aquasec.trivy.javadb.layer.v1.tar+gzip
+```
+
+If the above container registry is `gitlab.example.com/trivy-java-db-mirror`, then the container scanning job should be configured in the following way:
+
+```yaml
+include:
+  - template: Security/Container-Scanning.gitlab-ci.yml
+
+container_scanning:
+  variables:
+    CS_TRIVY_JAVA_DB: gitlab.example.com/trivy-java-db-mirror:1
+```
+
 ## Running the standalone container scanning tool
 
 It's possible to run the [GitLab container scanning tool](https://gitlab.com/gitlab-org/security-products/analyzers/container-scanning)
diff --git a/spec/models/ml/candidate_spec.rb b/spec/models/ml/candidate_spec.rb
index fa19b723ee2..d5b71e2c3f7 100644
--- a/spec/models/ml/candidate_spec.rb
+++ b/spec/models/ml/candidate_spec.rb
@@ -16,6 +16,7 @@ RSpec.describe Ml::Candidate, factory_default: :keep, feature_category: :mlops d
     it { is_expected.to belong_to(:user) }
     it { is_expected.to belong_to(:package) }
     it { is_expected.to belong_to(:ci_build).class_name('Ci::Build') }
+    it { is_expected.to belong_to(:model_version).class_name('Ml::ModelVersion') }
     it { is_expected.to have_many(:params) }
     it { is_expected.to have_many(:metrics) }
     it { is_expected.to have_many(:metadata) }
@@ -35,6 +36,45 @@ RSpec.describe Ml::Candidate, factory_default: :keep, feature_category: :mlops d
     it { expect(described_class.new.eid).to be_present }
   end
 
+  describe 'validation' do
+    let_it_be(:model) { create(:ml_models, project: candidate.project) }
+    let_it_be(:model_version1) { create(:ml_model_versions, model: model) }
+    let_it_be(:model_version2) { create(:ml_model_versions, model: model) }
+    let_it_be(:validation_candidate) do
+      create(:ml_candidates, model_version: model_version1, project: candidate.project)
+    end
+
+    let(:params) do
+      {
+        model_version: nil
+      }
+    end
+
+    subject(:errors) do
+      candidate = described_class.new(**params)
+      candidate.validate
+      candidate.errors
+    end
+
+    describe 'model_version' do
+      context 'when model_version is nil' do
+        it { expect(errors).not_to include(:model_version_id) }
+      end
+
+      context 'when no other candidate is associated to the model_version' do
+        let(:params) { { model_version: model_version2 } }
+
+        it { expect(errors).not_to include(:model_version_id) }
+      end
+
+      context 'when another candidate has model_version_id' do
+        let(:params) { { model_version: validation_candidate.model_version } }
+
+        it { expect(errors).to include(:model_version_id) }
+      end
+    end
+  end
+
   describe '.destroy' do
     let_it_be(:candidate_to_destroy) do
       create(:ml_candidates, :with_metrics_and_params, :with_metadata, :with_artifact)
diff --git a/spec/models/ml/model_version_spec.rb b/spec/models/ml/model_version_spec.rb
index 8f17d53787f..bfe53177a8a 100644
--- a/spec/models/ml/model_version_spec.rb
+++ b/spec/models/ml/model_version_spec.rb
@@ -18,6 +18,7 @@ RSpec.describe Ml::ModelVersion, feature_category: :mlops do
     it { is_expected.to belong_to(:project) }
     it { is_expected.to belong_to(:model) }
     it { is_expected.to belong_to(:package).class_name('Packages::MlModel::Package') }
+    it { is_expected.to have_one(:candidate).class_name('Ml::Candidate') }
   end
 
   describe 'validation' do
diff --git a/spec/services/ml/create_candidate_service_spec.rb b/spec/services/ml/create_candidate_service_spec.rb
new file mode 100644
index 00000000000..fb3456b0bcc
--- /dev/null
+++ b/spec/services/ml/create_candidate_service_spec.rb
@@ -0,0 +1,57 @@
+# frozen_string_literal: true
+
+require 'spec_helper'
+
+RSpec.describe ::Ml::CreateCandidateService, feature_category: :mlops do
+  describe '#execute' do
+    let_it_be(:model_version) { create(:ml_model_versions) }
+    let_it_be(:experiment) { create(:ml_experiments, project: model_version.project) }
+
+    let(:params) { {} }
+
+    subject(:candidate) { described_class.new(experiment, params).execute }
+
+    context 'with default parameters' do
+      it 'creates a candidate' do
+        expect { candidate }.to change { experiment.candidates.count }.by(1)
+      end
+
+      it 'gives a fake name' do
+        expect(candidate.name).to match(/[a-z]+-[a-z]+-[a-z]+-\d+/)
+      end
+
+      it 'sets the correct values', :aggregate_failures do
+        expect(candidate.start_time).to eq(0)
+        expect(candidate.experiment).to be(experiment)
+        expect(candidate.project).to be(experiment.project)
+        expect(candidate.user).to be_nil
+      end
+    end
+
+    context 'when parameters are passed' do
+      let(:params) do
+        {
+          start_time: 1234,
+          name: 'candidate_name',
+          model_version: model_version,
+          user: experiment.user
+        }
+      end
+
+      context 'with default parameters' do
+        it 'creates a candidate' do
+          expect { candidate }.to change { experiment.candidates.count }.by(1)
+        end
+
+        it 'sets the correct values', :aggregate_failures do
+          expect(candidate.start_time).to eq(1234)
+          expect(candidate.experiment).to be(experiment)
+          expect(candidate.project).to be(experiment.project)
+          expect(candidate.user).to be(experiment.user)
+          expect(candidate.name).to eq('candidate_name')
+          expect(candidate.model_version_id).to eq(model_version.id)
+        end
+      end
+    end
+  end
+end
diff --git a/spec/services/ml/find_or_create_model_version_service_spec.rb b/spec/services/ml/find_or_create_model_version_service_spec.rb
index 382d0037dbb..80391ffcadb 100644
--- a/spec/services/ml/find_or_create_model_version_service_spec.rb
+++ b/spec/services/ml/find_or_create_model_version_service_spec.rb
@@ -28,6 +28,7 @@ RSpec.describe ::Ml::FindOrCreateModelVersionService, feature_category: :mlops d
 
       it 'returns existing model version', :aggregate_failures do
         expect { model_version }.to change { Ml::ModelVersion.count }.by(0)
+        expect { model_version }.to change { Ml::Candidate.count }.by(0)
         expect(model_version).to eq(existing_version)
       end
     end
@@ -41,11 +42,12 @@ RSpec.describe ::Ml::FindOrCreateModelVersionService, feature_category: :mlops d
       let(:package) { create(:ml_model_package, project: project, name: name, version: version) }
 
       it 'creates a new model version', :aggregate_failures do
-        expect { model_version }.to change { Ml::ModelVersion.count }
+        expect { model_version }.to change { Ml::ModelVersion.count }.by(1).and change { Ml::Candidate.count }.by(1)
 
         expect(model_version.name).to eq(name)
         expect(model_version.version).to eq(version)
         expect(model_version.package).to eq(package)
+        expect(model_version.candidate.model_version_id).to eq(model_version.id)
         expect(model_version.description).to eq(description)
       end
     end
diff --git a/spec/tooling/danger/rubocop_inline_disable_suggestion_spec.rb b/spec/tooling/danger/rubocop_inline_disable_suggestion_spec.rb
index 950c313b9b9..6b9ff667564 100644
--- a/spec/tooling/danger/rubocop_inline_disable_suggestion_spec.rb
+++ b/spec/tooling/danger/rubocop_inline_disable_suggestion_spec.rb
@@ -15,10 +15,15 @@ RSpec.describe Tooling::Danger::RubocopInlineDisableSuggestion, feature_category
 
   let(:template) do
     <<~SUGGESTION_MARKDOWN.chomp
+    ```suggestion
+    %<suggested_line>s
+    ```
 
     Consider removing this inline disabling and adhering to the rubocop rule.
-    If that isn't possible, please provide context as a reply for reviewers.
-    See [rubocop best practices](https://docs.gitlab.com/ee/development/rubocop_development_guide.html).
+
+    If that isn't possible, please provide the reason as a code comment in the
+    same line where the rule is disabled separated by ` -- `.
+    See [rubocop best practices](https://docs.gitlab.com/ee/development/rubocop_development_guide.html#disabling-rules-inline).
 
     ----
 
@@ -77,6 +82,9 @@ RSpec.describe Tooling::Danger::RubocopInlineDisableSuggestion, feature_category
 
       def show_my_new_dot?(project, namespace)
         return false unless ::Gitlab.com? # rubocop: todo Gitlab/AvoidGitlabInstanceChecks -- Reason for disabling
+        thatsfine = "".dup # rubocop:disable Lint/UselessAssignment,Performance/UnfreezeString -- That's OK
+        me = "".dup # rubocop:disable Lint/UselessAssignment,Performance/UnfreezeString
+        test = "".dup # rubocop:disable Lint/UselessAssignment, Performance/UnfreezeString
         return false if notification_dot_acknowledged?
 
         show_out_of_pipeline_minutes_notification?(project, namespace)
@@ -102,6 +110,8 @@ RSpec.describe Tooling::Danger::RubocopInlineDisableSuggestion, feature_category
       +  return false unless ::Gitlab.com? # rubocop:todo Gitlab/AvoidGitlabInstanceChecks
       +  return false unless ::Gitlab.com? # rubocop: todo Gitlab/AvoidGitlabInstanceChecks
       +  return false unless ::Gitlab.com? # rubocop: todo Gitlab/AvoidGitlabInstanceChecks -- Reason for disabling
+      +  me = "".dup # rubocop:disable Lint/UselessAssignment,Performance/UnfreezeString
+      +  test = "".dup # rubocop:disable Lint/UselessAssignment, Performance/UnfreezeString
       +  return false unless ::Gitlab.com? # rubocop: todo Gitlab/AvoidGitlabInstanceChecks --
     DIFF
   end
@@ -119,8 +129,12 @@ RSpec.describe Tooling::Danger::RubocopInlineDisableSuggestion, feature_category
   end
 
   it 'adds comments at the correct lines', :aggregate_failures do
-    [3, 7, 13, 20, 27, 34, 41, 55].each do |line_number|
-      expect(rubocop).to receive(:markdown).with(template, file: filename, line: line_number)
+    [3, 7, 13, 20, 27, 34, 41, 50, 51, 58].each do |line_number|
+      existing_line = file_lines[line_number - 1].sub(/ --\s*$/, '')
+      suggested_line = "#{existing_line} -- TODO: Reason why the rule must be disabled"
+      comment = format(template, suggested_line: suggested_line)
+
+      expect(rubocop).to receive(:markdown).with(comment, file: filename, line: line_number)
     end
 
     rubocop.add_suggestions_for(filename)
diff --git a/tooling/danger/rubocop_inline_disable_suggestion.rb b/tooling/danger/rubocop_inline_disable_suggestion.rb
index ad481ef1334..589816a5937 100644
--- a/tooling/danger/rubocop_inline_disable_suggestion.rb
+++ b/tooling/danger/rubocop_inline_disable_suggestion.rb
@@ -5,13 +5,15 @@ require_relative 'suggestion'
 module Tooling
   module Danger
     class RubocopInlineDisableSuggestion < Suggestion
-      MATCH = /^\+.*#\s*rubocop\s*:\s*(?:disable|todo)\s+(?!.*\s--\s\S)/
-      REPLACEMENT = nil
+      MATCH = %r{^(?<line>.*#\s*rubocop\s*:\s*(?:disable|todo)\s+(?:[\w/]+(?:\s*,\s*[\w/]+)*))\s*(?!.*\s*--\s\S).*}
+      REPLACEMENT = '\k<line> -- TODO: Reason why the rule must be disabled'
 
       SUGGESTION = <<~MESSAGE_MARKDOWN
         Consider removing this inline disabling and adhering to the rubocop rule.
-        If that isn't possible, please provide context as a reply for reviewers.
-        See [rubocop best practices](https://docs.gitlab.com/ee/development/rubocop_development_guide.html).
+
+        If that isn't possible, please provide the reason as a code comment in the
+        same line where the rule is disabled separated by ` -- `.
+        See [rubocop best practices](https://docs.gitlab.com/ee/development/rubocop_development_guide.html#disabling-rules-inline).
 
         ----