Merge shuffle score pods logic

[learner0810]

What type of PR is this?
/kind cleanup

What this PR does / why we need it:

Merge shuffle score pods logic

Which issue(s) this PR fixes:

Fixes #

Does this PR introduce a user-facing change?:

[netlify]

✅ Deploy Preview for gateway-api-inference-extension ready!

Name	Link
🔨 Latest commit	6ac1fe4
🔍 Latest deploy log	https://app.netlify.com/projects/gateway-api-inference-extension/deploys/68bfe268baae590007b1f4dc
😎 Deploy Preview	https://deploy-preview-1552--gateway-api-inference-extension.netlify.app
📱 Preview on mobile	Toggle QR Code... Use your smartphone camera to open QR code link.

---

To edit notification comments on pull requests, go to your Netlify project configuration.

[k8s-ci-robot]

Hi @learner0810. Thanks for your PR.

I'm waiting for a kubernetes-sigs member to verify that this patch is reasonable to test. If it is, they should reply with /ok-to-test on its own line. Until that is done, I will not automatically test new commits in this PR, but the usual testing commands by org members will still work. Regular contributors should join the org to skip this step.

Once the patch is verified, the new status will be reflected by the ok-to-test label.

I understand the commands that are listed here.

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository.

[elevran]

/ok-to-test

[kfswain]

/lgtm
/approve

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: kfswain, learner0810

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• OWNERS [kfswain]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[elevran]

have you compared using sync.Pool vs creating a new random source every time?
I don't expect much latency difference but a sync.Pool would reduce GC pressure as these are created and collected per request.

[learner0810]

func shuffleScoredPods(scoredPods []*types.ScoredPod) {
	// Rand package is not safe for concurrent use, so we create a new instance.
	// Source: https://pkg.go.dev/math/rand/v2#pkg-overview
	randomGenerator := rand.New(rand.NewPCG(uint64(time.Now().UnixNano()), 0))

	// Shuffle in-place
	randomGenerator.Shuffle(len(scoredPods), func(i, j int) {
		scoredPods[i], scoredPods[j] = scoredPods[j], scoredPods[i]
	})
}

var randPool = sync.Pool{
	New: func() interface{} {
		return rand.New(rand.NewPCG(uint64(time.Now().UnixNano()), 0))
	},
}

func shuffleScoredPodsWithPool(scoredPods []*types.ScoredPod) {
	randomGenerator := randPool.Get().(*rand.Rand)
	defer randPool.Put(randomGenerator) 

	// Shuffle in-place
	randomGenerator.Shuffle(len(scoredPods), func(i, j int) {
		scoredPods[i], scoredPods[j] = scoredPods[j], scoredPods[i]
	})
}

func createTestScoredPods(count int) []*types.ScoredPod {
	pods := make([]*types.ScoredPod, count)
	for i := 0; i < count; i++ {
		pods[i] = &types.ScoredPod{
			Pod: &types.PodMetrics{
				Pod: &backend.Pod{
					NamespacedName: k8stypes.NamespacedName{
						Name:      "pod-" + strconv.Itoa(i),
						Namespace: "namespace-" + strconv.Itoa(i),
					},
				},
			},
			Score: float64(i),
		}
	}
	return pods
}

func BenchmarkShuffleScoredPods(b *testing.B) {
	testPods := createTestScoredPods(100)
	b.ResetTimer()

	for i := 0; i < b.N; i++ {
		podsCopy := make([]*types.ScoredPod, len(testPods))
		copy(podsCopy, testPods)

		shuffleScoredPods(podsCopy)
	}
}

func BenchmarkShuffleScoredPodsWithPool(b *testing.B) {
	testPods := createTestScoredPods(100)
	b.ResetTimer()

	for i := 0; i < b.N; i++ {
		podsCopy := make([]*types.ScoredPod, len(testPods))
		copy(podsCopy, testPods)

		shuffleScoredPodsWithPool(podsCopy)
	}
}

func BenchmarkShuffleScoredPods_Small(b *testing.B) {
	testPods := createTestScoredPods(10)
	b.ResetTimer()

	for i := 0; i < b.N; i++ {
		podsCopy := make([]*types.ScoredPod, len(testPods))
		copy(podsCopy, testPods)

		shuffleScoredPods(podsCopy)
	}
}

func BenchmarkShuffleScoredPodsWithPool_Small(b *testing.B) {
	testPods := createTestScoredPods(10)
	b.ResetTimer()

	for i := 0; i < b.N; i++ {
		podsCopy := make([]*types.ScoredPod, len(testPods))
		copy(podsCopy, testPods)

		shuffleScoredPodsWithPool(podsCopy)
	}
}

func BenchmarkShuffleScoredPods_Large(b *testing.B) {
	testPods := createTestScoredPods(1000)
	b.ResetTimer()

	for i := 0; i < b.N; i++ {
		podsCopy := make([]*types.ScoredPod, len(testPods))
		copy(podsCopy, testPods)

		shuffleScoredPods(podsCopy)
	}
}

func BenchmarkShuffleScoredPodsWithPool_Large(b *testing.B) {
	testPods := createTestScoredPods(1000)
	b.ResetTimer()

	for i := 0; i < b.N; i++ {
		podsCopy := make([]*types.ScoredPod, len(testPods))
		copy(podsCopy, testPods)

		shuffleScoredPodsWithPool(podsCopy)
	}
}

I have tested it locally and the test result is as follows. If there is any problem, please point it out. If the test results are accurate, do you think it needs to be optimized?

host@hostdeMacBook-Pro picker % go test -bench=. -benchmem
goos: darwin
goarch: amd64
pkg: sigs.k8s.io/gateway-api-inference-extension/pkg/epp/scheduling/framework/plugins/picker
cpu: Intel(R) Core(TM) i5-1038NG7 CPU @ 2.00GHz
BenchmarkShuffleScoredPods-8                      911576              1245 ns/op             912 B/op          2 allocs/op
BenchmarkShuffleScoredPodsWithPool-8             1000000              1130 ns/op             896 B/op          1 allocs/op
BenchmarkShuffleScoredPods_Small-8               4351080               247.4 ns/op            96 B/op          2 allocs/op
BenchmarkShuffleScoredPodsWithPool_Small-8       7934428               139.8 ns/op            80 B/op          1 allocs/op
BenchmarkShuffleScoredPods_Large-8                115184             10405 ns/op            8208 B/op          2 allocs/op
BenchmarkShuffleScoredPodsWithPool_Large-8        115098             10628 ns/op            8195 B/op          1 allocs/op
PASS
ok      sigs.k8s.io/gateway-api-inference-extension/pkg/epp/scheduling/framework/plugins/picker 7.652s

[elevran]

please keep import groups separated and not combine controller-runtime (3rd-party/ external) with gateway-api-inference-extension (local/internal)

[elevran]

ditto