Fix: Move replica validation logic to right place. #4307

kash2104 · 2025-12-23T13:21:58Z

Why are these changes needed?

These changes are done to check the validation logic before rayCluster pod creation. It moves the replica validation logic as well as removes the redundant tests. Along with this, unit test are added since we moved the logic from utils.go to validation.go.

Related issue number

Closes #4101

Checks

I've made sure the tests are passing.
Testing Strategy
- Unit tests
- Manual tests
- This PR is not tested :(

JiangJiaWei1103 · 2025-12-24T08:09:54Z

Thanks for your effort.

cc @Future-Outlier
This PR duplicates with #4116, which seemed inactive now. Should we close #4116 and work on this one?

kash2104 · 2025-12-24T09:27:15Z

I have gone through #4116 but it was inactive and no changes were being made after the merging of volcano pr. Many changes are required since volcano pr's merging, that's why I opened up this PR.

JiangJiaWei1103 · 2025-12-24T14:07:43Z

Gotcha. Let's wait for maintainers' reply. I'll help review, thank you.

kash2104 · 2026-01-01T04:40:48Z

@Future-Outlier Just wanted to follow up - do you prefer closing #4116 and continuing the work in this PR?

Future-Outlier · 2026-01-01T04:45:36Z

@Future-Outlier Just wanted to follow up - do you prefer closing #4116 and continuing the work in this PR?

Hi, @kash2104 I just left a comment to #4116 and ask her if she have time to finish the work.
if after 2 weeks she haven't replied, you can take over it

Future-Outlier

cc @400Ping @justinyeh1995 @machichima to review

Future-Outlier · 2026-01-14T06:27:02Z

cursor review

ray-operator/controllers/ray/utils/validation.go

cursor

Cursor Bugbot has reviewed your changes and found 1 potential issue.

ray-operator/controllers/ray/utils/validation.go

machichima · 2026-01-18T06:46:09Z

ray-operator/controllers/ray/utils/validation.go

+			if *workerGroup.MinReplicas > *workerGroup.MaxReplicas {
+				return fmt.Errorf("worker group %s has minReplicas %d greater than maxReplicas %d", workerGroup.GroupName, *workerGroup.MinReplicas, *workerGroup.MaxReplicas)
+			}
+		}


Should we also check if workerGroup.Replicas lies within the workerGroup.MinReplicas and workerGroup.MaxReplicas`?

@machichima Here in thevalidation.go, we are just checking whether the values for min and max replicas aren't incorrect or impossible before pod creation happens but the actual logic of number of replicas is moved to util.go

So I think that here in validation.go, we won't be needing this check.

machichima · 2026-01-18T06:55:01Z

ray-operator/controllers/ray/utils/util.go

+		if nodeGroup.MinReplicas != nil {
+			minReplicas = *nodeGroup.MinReplicas
+		}


Maybe we can use ptr.Deref, which is used in other places as well

kuberay/ray-operator/controllers/ray/utils/util.go

Lines 830 to 831 in 270a143

ptr.Deref(rayServiceInstance.Status.ActiveServiceStatus.TargetCapacity, -1) == 0 &&

ptr.Deref(rayServiceInstance.Status.PendingServiceStatus.TrafficRoutedPercent, -1) == 100

1. Remove validation logic from GetWorkerGroupDesiredReplicas (utils.go) and add this logic to ValidateRayClusterSpec (validation.go). 2. Remove unnecessary tests from TestGetWorkerGroupDesiredReplicas. 3. Remove the unused ctx.

This is added since we moved the validation logic.

ray-operator/controllers/ray/utils/util.go

machichima · 2026-01-19T11:29:57Z

ray-operator/controllers/ray/utils/util_test.go

-	// Test 3: `WorkerGroupSpec.Replicas` is not nil but is more than maxReplicas.
-	replicas = int32(6)
-	workerGroupSpec.Replicas = &replicas
-	assert.Equal(t, GetWorkerGroupDesiredReplicas(ctx, workerGroupSpec), maxReplicas)
-
-	// Test 4: `WorkerGroupSpec.Replicas` is not nil but is less than minReplicas.
-	replicas = int32(0)
-	workerGroupSpec.Replicas = &replicas
-	assert.Equal(t, GetWorkerGroupDesiredReplicas(ctx, workerGroupSpec), minReplicas)


Should we keep those checks? I think this part is not moved to validation

The part that we have moved to the validation keeps in check that the desired state of the replicas are within the correct bounds or not after the pods have been created. So these tests were checking that logic by using the GetWorkerGroupDesiredReplicas and we have moved this logic to validation.go and thus have added the tests for them in validation_test.go as well.

Thus these checks have removed from util_test.go.

Please correct me if I'm wrong, I think in validation we only check if maxReplicas and minReplicas are set correctly (comment: #4307 (comment)). The test here is for checking if the replicas will be clamped to max or min replica, so I think we should include them?

machichima · 2026-01-19T11:59:44Z

ray-operator/controllers/ray/utils/validation_test.go

+		{
+			name: "replicas smaller than minReplicas when autoscaling disabled",
+			spec: func() rayv1.RayClusterSpec {
+				s := createSpec()
+				s.WorkerGroupSpecs = []rayv1.WorkerGroupSpec{
+					{
+						GroupName:   "worker-group-3",
+						Template:    podTemplateSpec(nil, nil),
+						Replicas:    ptr.To(int32(1)),
+						MinReplicas: ptr.To(int32(2)),
+						MaxReplicas: ptr.To(int32(5)),
+					},
+				}
+				return s
+			}(),
+			expectError: false,
+		},


I think this will not raise error because we did the clamp it to minReplicas in util.go? Could we please add comment or update the test name to briefly explain why this will not raise error? As usually we would expect error being raised in this case

kash2104 requested review from MortalHappiness, andrewsykim, kevin85421 and rueian as code owners December 23, 2025 13:21

kash2104 force-pushed the issue-4101 branch 3 times, most recently from d1d7b57 to eb654d7 Compare December 24, 2025 06:09

kash2104 force-pushed the issue-4101 branch from eb654d7 to b00b0cb Compare December 24, 2025 09:43

JiangJiaWei1103 moved this to Backlog in My Kuberay & Ray Dec 24, 2025

JiangJiaWei1103 added this to My Kuberay & Ray Dec 24, 2025

kash2104 force-pushed the issue-4101 branch 3 times, most recently from 529b021 to bc5decd Compare January 1, 2026 17:00

Future-Outlier added this to @Future-Outlier's kuberay project Jan 12, 2026

Future-Outlier moved this to In Progress in @Future-Outlier's kuberay project Jan 12, 2026

Future-Outlier moved this from In Progress to to review in @Future-Outlier's kuberay project Jan 12, 2026

Future-Outlier reviewed Jan 12, 2026

View reviewed changes

cursor bot reviewed Jan 14, 2026

View reviewed changes

ray-operator/controllers/ray/utils/validation.go Show resolved Hide resolved

kash2104 force-pushed the issue-4101 branch from bc5decd to c36d0b4 Compare January 17, 2026 13:31

cursor bot reviewed Jan 17, 2026

View reviewed changes

ray-operator/controllers/ray/utils/validation.go Show resolved Hide resolved

kash2104 force-pushed the issue-4101 branch from c36d0b4 to 270a143 Compare January 17, 2026 14:12

machichima reviewed Jan 18, 2026

View reviewed changes

Fix: Replica validation logic in right place.

17f7099

1. Remove validation logic from GetWorkerGroupDesiredReplicas (utils.go) and add this logic to ValidateRayClusterSpec (validation.go). 2. Remove unnecessary tests from TestGetWorkerGroupDesiredReplicas. 3. Remove the unused ctx.

kash2104 force-pushed the issue-4101 branch from 270a143 to ad6fd8a Compare January 18, 2026 18:38

Test: Add unit test for replica validation.

7d73dd1

This is added since we moved the validation logic.

kash2104 force-pushed the issue-4101 branch from ad6fd8a to 7d73dd1 Compare January 19, 2026 03:47

Future-Outlier moved this from to review to In Progress in @Future-Outlier's kuberay project Jan 19, 2026

machichima reviewed Jan 19, 2026

View reviewed changes

	ptr.Deref(rayServiceInstance.Status.ActiveServiceStatus.TargetCapacity, -1) == 0 &&
	ptr.Deref(rayServiceInstance.Status.PendingServiceStatus.TrafficRoutedPercent, -1) == 100

Fix: Move replica validation logic to right place. #4307

Are you sure you want to change the base?

Fix: Move replica validation logic to right place. #4307

Uh oh!

Conversation

kash2104 commented Dec 23, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Why are these changes needed?

Related issue number

Checks

Uh oh!

JiangJiaWei1103 commented Dec 24, 2025

Uh oh!

kash2104 commented Dec 24, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

JiangJiaWei1103 commented Dec 24, 2025

Uh oh!

kash2104 commented Jan 1, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Future-Outlier commented Jan 1, 2026

Uh oh!

Future-Outlier left a comment

Choose a reason for hiding this comment

Uh oh!

Future-Outlier commented Jan 14, 2026

Uh oh!

Uh oh!

cursor bot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

machichima Jan 18, 2026

Choose a reason for hiding this comment

Uh oh!

kash2104 Jan 18, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

machichima Jan 18, 2026

Choose a reason for hiding this comment

Uh oh!

Uh oh!

machichima Jan 19, 2026

Choose a reason for hiding this comment

Uh oh!

kash2104 Jan 19, 2026

Choose a reason for hiding this comment

Uh oh!

machichima Jan 19, 2026

Choose a reason for hiding this comment

Uh oh!

machichima Jan 19, 2026

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

kash2104 commented Dec 23, 2025 •

edited

Loading

kash2104 commented Dec 24, 2025 •

edited

Loading

kash2104 commented Jan 1, 2026 •

edited

Loading

kash2104 Jan 18, 2026 •

edited

Loading