Expire project caches once per push instead of once per ref

Previously `ProjectCacheWorker` would be scheduled once per ref, which
would generate unnecessary I/O and load on Sidekiq, especially if many
tags or branches were pushed at once. `ProjectCacheWorker` would expire
three items:

1. Repository size: This only needs to be updated once per push.
2. Commit count: This only needs to be updated if the default branch
   is updated.
3. Project method caches: This only needs to be updated if the default
   branch changes, but only if certain files change (e.g. README,
   CHANGELOG, etc.).

Because the third item requires looking at the actual changes in the
commit deltas, we schedule one `ProjectCacheWorker` to handle the first
two cases, and schedule a separate `ProjectCacheWorker` for the third
case if it is needed. As a result, this brings down the number of
`ProjectCacheWorker` jobs from N to 2.

Closes https://gitlab.com/gitlab-org/gitlab-ce/issues/52046

5 files changed, 144 insertions, 12 deletions

diff --git a/spec/lib/gitlab/git_post_receive_spec.rb b/spec/lib/gitlab/git_post_receive_spec.rb
index 4c20d945585..f0df3794e29 100644
--- a/spec/lib/gitlab/git_post_receive_spec.rb
+++ b/spec/lib/gitlab/git_post_receive_spec.rb
@@ -3,7 +3,7 @@
 require 'spec_helper'
 
 describe ::Gitlab::GitPostReceive do
-  let(:project) { create(:project) }
+  set(:project) { create(:project, :repository) }
 
   subject { described_class.new(project, "project-#{project.id}", changes.dup, {}) }
 
@@ -92,4 +92,47 @@ describe ::Gitlab::GitPostReceive do
       end
     end
   end
+
+  describe '#includes_default_branch?' do
+    context 'with no default branch' do
+      let(:changes) do
+        <<~EOF
+          654321 210987 refs/heads/test1
+          654322 210986 refs/tags/#{project.default_branch}
+          654323 210985 refs/heads/test3
+        EOF
+      end
+
+      it 'returns false' do
+        expect(subject.includes_default_branch?).to be_falsey
+      end
+    end
+
+    context 'with a project with no default branch' do
+      let(:changes) do
+        <<~EOF
+          654321 210987 refs/heads/test1
+        EOF
+      end
+
+      it 'returns true' do
+        expect(project).to receive(:default_branch).and_return(nil)
+        expect(subject.includes_default_branch?).to be_truthy
+      end
+    end
+
+    context 'with default branch' do
+      let(:changes) do
+        <<~EOF
+          654322 210986 refs/heads/test1
+          654321 210987 refs/tags/test2
+          654323 210985 refs/heads/#{project.default_branch}
+        EOF
+      end
+
+      it 'returns true' do
+        expect(subject.includes_default_branch?).to be_truthy
+      end
+    end
+  end
 end
diff --git a/spec/models/repository_spec.rb b/spec/models/repository_spec.rb
index e68de2e73a8..419e1dc2459 100644
--- a/spec/models/repository_spec.rb
+++ b/spec/models/repository_spec.rb
@@ -1815,22 +1815,36 @@ describe Repository do
   end
 
   describe '#after_create' do
+    it 'calls expire_status_cache' do
+      expect(repository).to receive(:expire_status_cache)
+
+      repository.after_create
+    end
+
+    it 'logs an event' do
+      expect(repository).to receive(:repository_event).with(:create_repository)
+
+      repository.after_create
+    end
+  end
+
+  describe '#expire_status_cache' do
     it 'flushes the exists cache' do
       expect(repository).to receive(:expire_exists_cache)
 
-      repository.after_create
+      repository.expire_status_cache
     end
 
     it 'flushes the root ref cache' do
       expect(repository).to receive(:expire_root_ref_cache)
 
-      repository.after_create
+      repository.expire_status_cache
     end
 
     it 'flushes the emptiness caches' do
       expect(repository).to receive(:expire_emptiness_caches)
 
-      repository.after_create
+      repository.expire_status_cache
     end
   end
 
diff --git a/spec/services/git/branch_hooks_service_spec.rb b/spec/services/git/branch_hooks_service_spec.rb
index 8af51848b7b..3929f51a0e2 100644
--- a/spec/services/git/branch_hooks_service_spec.rb
+++ b/spec/services/git/branch_hooks_service_spec.rb
@@ -158,9 +158,13 @@ describe Git::BranchHooksService do
     let(:blank_sha) { Gitlab::Git::BLANK_SHA }
 
     def clears_cache(extended: [])
-      expect(ProjectCacheWorker)
-        .to receive(:perform_async)
-        .with(project.id, extended, %i[commit_count repository_size])
+      expect(service).to receive(:invalidated_file_types).and_return(extended)
+
+      if extended.present?
+        expect(ProjectCacheWorker)
+          .to receive(:perform_async)
+          .with(project.id, extended, [], false)
+      end
 
       service.execute
     end
diff --git a/spec/workers/post_receive_spec.rb b/spec/workers/post_receive_spec.rb
index 3b69b81f12e..c8a0c22b0e8 100644
--- a/spec/workers/post_receive_spec.rb
+++ b/spec/workers/post_receive_spec.rb
@@ -37,6 +37,29 @@ describe PostReceive do
   end
 
   describe "#process_project_changes" do
+    context 'with an empty project' do
+      let(:empty_project) { create(:project, :empty_repo) }
+      let(:changes) { "123456 789012 refs/heads/tést1\n" }
+
+      before do
+        allow_any_instance_of(Gitlab::GitPostReceive).to receive(:identify).and_return(empty_project.owner)
+        allow(Gitlab::GlRepository).to receive(:parse).and_return([empty_project, Gitlab::GlRepository::PROJECT])
+      end
+
+      it 'expire the status cache' do
+        expect(empty_project.repository).to receive(:expire_status_cache)
+
+        perform
+      end
+
+      it 'schedules a cache update for commit count and size' do
+        expect(ProjectCacheWorker).to receive(:perform_async)
+                                        .with(empty_project.id, [], [:repository_size, :commit_count], true)
+
+        perform
+      end
+    end
+
     context 'empty changes' do
       it "does not call any PushService but runs after project hooks" do
         expect(Git::BranchPushService).not_to receive(:new)
@@ -67,15 +90,22 @@ describe PostReceive do
       context "branches" do
         let(:changes) do
           <<~EOF
-            '123456 789012 refs/heads/tést1'
-            '123456 789012 refs/heads/tést2'
+            123456 789012 refs/heads/tést1
+            123456 789012 refs/heads/tést2
           EOF
         end
 
         it 'expires the branches cache' do
           expect(project.repository).to receive(:expire_branches_cache).once
 
-          described_class.new.perform(gl_repository, key_id, base64_changes)
+          perform
+        end
+
+        it 'expires the status cache' do
+          expect(project).to receive(:empty_repo?).and_return(true)
+          expect(project.repository).to receive(:expire_status_cache)
+
+          perform
         end
 
         it 'calls Git::BranchPushService' do
@@ -87,6 +117,30 @@ describe PostReceive do
 
           perform
         end
+
+        it 'schedules a cache update for repository size only' do
+          expect(ProjectCacheWorker).to receive(:perform_async)
+                                          .with(project.id, [], [:repository_size], true)
+
+          perform
+        end
+
+        context 'with a default branch' do
+          let(:changes) do
+            <<~EOF
+              123456 789012 refs/heads/tést1
+              123456 789012 refs/heads/tést2
+              678912 123455 refs/heads/#{project.default_branch}
+            EOF
+          end
+
+          it 'schedules a cache update for commit count and size' do
+            expect(ProjectCacheWorker).to receive(:perform_async)
+                                            .with(project.id, [], [:repository_size, :commit_count], true)
+
+            perform
+          end
+        end
       end
 
       context "tags" do
@@ -107,7 +161,7 @@ describe PostReceive do
         it 'does not expire branches cache' do
           expect(project.repository).not_to receive(:expire_branches_cache)
 
-          described_class.new.perform(gl_repository, key_id, base64_changes)
+          perform
         end
 
         it "only invalidates tags once" do
@@ -115,7 +169,7 @@ describe PostReceive do
           expect(project.repository).to receive(:expire_caches_for_tags).once.and_call_original
           expect(project.repository).to receive(:expire_tags_cache).once.and_call_original
 
-          described_class.new.perform(gl_repository, key_id, base64_changes)
+          perform
         end
 
         it "calls Git::TagPushService" do
@@ -129,6 +183,13 @@ describe PostReceive do
 
           perform
         end
+
+        it 'schedules a single ProjectCacheWorker update' do
+          expect(ProjectCacheWorker).to receive(:perform_async)
+                                          .with(project.id, [], [:repository_size], true)
+
+          perform
+        end
       end
 
       context "merge-requests" do
diff --git a/spec/workers/project_cache_worker_spec.rb b/spec/workers/project_cache_worker_spec.rb
index edc55920b8e..7f3c4881b89 100644
--- a/spec/workers/project_cache_worker_spec.rb
+++ b/spec/workers/project_cache_worker_spec.rb
@@ -49,6 +49,16 @@ describe ProjectCacheWorker do
         worker.perform(project.id, %w(readme))
       end
 
+      context 'with statistics disabled' do
+        let(:statistics) { [] }
+
+        it 'does not update the project statistics' do
+          expect(worker).not_to receive(:update_statistics)
+
+          worker.perform(project.id, [], [], false)
+        end
+      end
+
       context 'with statistics' do
         let(:statistics) { %w(repository_size) }