From 5ac9ec00af3b745ed012a52e72065e18abbcda50 Mon Sep 17 00:00:00 2001
From: Marek Siarkowicz <siarkowicz@google.com>
Date: Wed, 22 Oct 2025 16:22:40 +0200
Subject: [PATCH] Benchmark SharedIndexInformer

Kubernetes-commit: 7a9d718cdf5e26f83ce45cd5b24f06ef7e678bea
---
 tools/cache/shared_informer_bench_test.go | 160 ++++++++++++++++++++++
 1 file changed, 160 insertions(+)
 create mode 100644 tools/cache/shared_informer_bench_test.go

diff --git a/tools/cache/shared_informer_bench_test.go b/tools/cache/shared_informer_bench_test.go
new file mode 100644
index 000000000..5ac06ae1f
--- /dev/null
+++ b/tools/cache/shared_informer_bench_test.go
@@ -0,0 +1,160 @@
+/*
+Copyright 2025 The Kubernetes Authors.
+
+Licensed under the Apache License, Version 2.0 (the "License");
+you may not use this file except in compliance with the License.
+You may obtain a copy of the License at
+
+    http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software
+distributed under the License is distributed on an "AS IS" BASIS,
+WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+See the License for the specific language governing permissions and
+limitations under the License.
+*/
+
+package cache
+
+import (
+	"context"
+	"fmt"
+	"sync"
+	"sync/atomic"
+	"testing"
+	"time"
+
+	corev1 "k8s.io/api/core/v1"
+	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
+	"k8s.io/apimachinery/pkg/labels"
+	"k8s.io/apimachinery/pkg/runtime"
+	"k8s.io/apimachinery/pkg/util/rand"
+	"k8s.io/apimachinery/pkg/watch"
+)
+
+var benchmarkNamespace = "default"
+
+func BenchmarkSharedIndexInformer(b *testing.B) {
+	for _, podCount := range []int{10_000, 20_000, 40_000, 80_000, 160_000} {
+		b.Run(fmt.Sprintf("podCount=%d", podCount), func(b *testing.B) {
+			pods := createPods(podCount, benchmarkNamespace)
+			for _, readers := range []int{0, 1, 10, 20, 40, 80} {
+				b.Run(fmt.Sprintf("readers=%d", readers), func(b *testing.B) {
+					watcher := watch.NewFakeWithChanSize(1, false)
+					informer, stop := setupSharedIndexInformer(watcher, pods)
+					defer stop()
+					queuedEvents := 10
+					benchmarkSharedIndexInformer(b, readers, watcher, informer, pods, queuedEvents)
+				})
+			}
+		})
+	}
+}
+
+func createPods(count int, namespace string) []corev1.Pod {
+	pods := []corev1.Pod{}
+	for i := 0; i < count; i++ {
+		pod := &corev1.Pod{
+			ObjectMeta: metav1.ObjectMeta{
+				Name:      rand.String(20),
+				Namespace: namespace,
+			},
+		}
+		pods = append(pods, *pod)
+	}
+	return pods
+}
+
+func setupSharedIndexInformer(watcher watch.Interface, pods []corev1.Pod) (SharedIndexInformer, func()) {
+	podInformer := NewSharedIndexInformerWithOptions(
+		&ListWatch{
+			ListWithContextFunc: func(ctx context.Context, options metav1.ListOptions) (runtime.Object, error) {
+				return &corev1.PodList{
+					Items: pods,
+				}, nil
+			},
+			WatchFuncWithContext: func(ctx context.Context, options metav1.ListOptions) (watch.Interface, error) {
+				return watcher, nil
+			},
+		},
+		&corev1.Pod{},
+		SharedIndexInformerOptions{
+			ResyncPeriod: time.Second * 60,
+			Indexers:     Indexers{NamespaceIndex: MetaNamespaceIndexFunc},
+		},
+	)
+
+	stop := make(chan struct{})
+	var wg sync.WaitGroup
+	wg.Add(1)
+	go func() {
+		defer wg.Done()
+		podInformer.Run(stop)
+	}()
+	for !podInformer.HasSynced() {
+		time.Sleep(time.Millisecond)
+	}
+	return podInformer, func() {
+		close(stop)
+		wg.Wait()
+	}
+}
+
+func benchmarkSharedIndexInformer(b *testing.B, readers int, watcher *watch.FakeWatcher, podInformer SharedIndexInformer, pods []corev1.Pod, queuedEvents int) {
+	var writes atomic.Int64
+	got := make(chan struct{}, queuedEvents)
+
+	_, err := podInformer.AddEventHandler(ResourceEventHandlerDetailedFuncs{
+		UpdateFunc: func(oldObj, newObj interface{}) {
+			writes.Add(1)
+			select {
+			case <-got:
+			default:
+			}
+		},
+	})
+	if err != nil {
+		b.Fatal(err)
+	}
+	pod := pods[rand.Intn(len(pods))]
+	watcher.Modify(&pod)
+	// Wait for modify to arrive to confirm initial pods where handled.
+	for writes.Load() == 0 {
+		time.Sleep(time.Millisecond)
+	}
+
+	var wg sync.WaitGroup
+	stop := make(chan struct{})
+	var reads atomic.Int64
+	for i := 0; i < readers; i++ {
+		wg.Add(1)
+		go func() {
+			defer wg.Done()
+			for {
+				select {
+				case <-stop:
+					return
+				default:
+				}
+				err := ListAllByNamespace(podInformer.GetIndexer(), benchmarkNamespace, labels.Everything(), func(obj interface{}) {
+				})
+				if err != nil {
+					panic(err)
+				}
+				reads.Add(1)
+			}
+		}()
+	}
+
+	writes.Store(0)
+	reads.Store(0)
+	for b.Loop() {
+		pod := pods[rand.Intn(len(pods))]
+		watcher.Modify(&pod)
+		got <- struct{}{}
+	}
+	b.ReportMetric(float64(reads.Load())/b.Elapsed().Seconds(), "reads/s")
+	b.ReportMetric(float64(writes.Load())/b.Elapsed().Seconds(), "writes/s")
+	close(stop)
+	wg.Wait()
+}