WaitForPodsReady: Reset the requeueState while reconciling #1838

tenzen-y · 2024-03-13T23:20:07Z

What type of PR is this?

/kind bug

What this PR does / why we need it:

As @alculquicondor mentioned #1821 (comment), mutating webhooks seems not to be possible to update spec and status in a single webhook call.

So, we must reset requeueState while reconciling instead of webhooks.

Which issue(s) this PR fixes:

Fixes #1821

Special notes for your reviewer:

Does this PR introduce a user-facing change?

WaitForPodsReady: Fix a bug that the requeueState isn't reset.

netlify · 2024-03-13T23:20:25Z

✅ Deploy Preview for kubernetes-sigs-kueue canceled.

Name	Link
🔨 Latest commit	`688f223`
🔍 Latest deploy log	https://app.netlify.com/sites/kubernetes-sigs-kueue/deploys/65f3242a7a18de0008579f8d

tenzen-y · 2024-03-13T23:20:34Z

/cherry-pick release-0.6

k8s-infra-cherrypick-robot · 2024-03-13T23:20:41Z

@tenzen-y: once the present PR merges, I will cherry-pick it on top of release-0.6 in a new PR and assign it to you.

In response to this:

/cherry-pick release-0.6

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/test-infra repository.

gabesaba · 2024-03-14T14:35:59Z

test/integration/scheduler/podsready/scheduler_test.go

@@ -248,7 +248,7 @@ var _ = ginkgo.Describe("SchedulerWithWaitForPodsReady", func() {
 			// To avoid flakiness, we don't verify if the workload has a QuotaReserved=false with pending reason here.
 		})

-		ginkgo.It("Should re-admit a timed out workload and deactivate a workload exceeded the re-queue count limit", func() {
+		ginkgo.It("Should re-admit a timed out workload and deactivate a workload exceeded the re-queue count limit. After that re-activating a workload", func() {


nit: s/activating/activate

may we consider a separate test for this, assuming that setting up the state isn't too difficult?

It is possible to divide this test into two tests, but we need to start another manager to configure the backoffLimitCount=1 or 0 like this:

kueue/test/integration/scheduler/podsready/scheduler_test.go

Line 201 in 975b5f2

requeuingBackoffLimitCount = ptr.To[int32](2)

It could increase integration test time. So, to avoid increasing time, I implement this test like the current one.

gabesaba · 2024-03-14T14:41:19Z

test/integration/scheduler/podsready/scheduler_test.go

+			}, util.Timeout, util.Interval).Should(gomega.Succeed(), "Reactivate inactive Workload")
+			gomega.Eventually(func(g gomega.Gomega) {
+				g.Expect(k8sClient.Get(ctx, client.ObjectKeyFromObject(prodWl), prodWl)).Should(gomega.Succeed())
+				g.Expect(prodWl.Status.RequeueState).Should(gomega.BeNil())


should we verify at the start of this sequence that the Requeue state is in a certain configuration (non-nil, or something more specific), in case the previous part of test changes?

Not relevant if this test is split out and this state is set explicitly.

We already verify that here:

kueue/test/integration/scheduler/podsready/scheduler_test.go

Line 273 in 975b5f2

util.ExpectWorkloadToHaveRequeueCount(ctx, k8sClient, client.ObjectKeyFromObject(prodWl), ptr.To[int32](2))

Does it make sense?

alculquicondor · 2024-03-14T15:31:10Z

test/integration/scheduler/podsready/scheduler_test.go

+				apimeta.SetStatusCondition(&prodWl.Status.Conditions, metav1.Condition{
+					Type:    kueue.WorkloadEvicted,
+					Status:  metav1.ConditionTrue,
+					Reason:  kueue.WorkloadEvictedByDeactivation,
+					Message: "evicted by Test",
+				})
+				g.Expect(k8sClient.Status().Update(ctx, prodWl)).Should(gomega.Succeed())


wouldn't the controller add this already?

Although the jobs controller adds this condition here:

kueue/pkg/controller/jobframework/reconciler.go

Line 423 in a96927d

workload.SetEvictedCondition(wl, kueue.WorkloadEvictedByDeactivation, "The workload is deactivated")

, we don't register the jobs controller to the manager.

kueue/test/integration/scheduler/podsready/suite_test.go

Lines 68 to 104 in 975b5f2

cfg := &config.Configuration{

WaitForPodsReady: &config.WaitForPodsReady{

Enable: true,

BlockAdmission: &blockAdmission,

Timeout: &metav1.Duration{Duration: value},

RequeuingStrategy: &config.RequeuingStrategy{

Timestamp: ptr.To(requeuingTimestamp),

BackoffLimitCount: requeuingBackoffLimitCount,

},

},

}

mgr.GetScheme().Default(cfg)

err := indexer.Setup(ctx, mgr.GetFieldIndexer())

gomega.Expect(err).NotTo(gomega.HaveOccurred())

cCache := cache.New(mgr.GetClient(), cache.WithPodsReadyTracking(cfg.WaitForPodsReady.Enable && cfg.WaitForPodsReady.BlockAdmission != nil && *cfg.WaitForPodsReady.BlockAdmission))

queues := queue.NewManager(

mgr.GetClient(), cCache,

queue.WithPodsReadyRequeuingTimestamp(requeuingTimestamp),

)

failedCtrl, err := core.SetupControllers(mgr, queues, cCache, cfg)

gomega.Expect(err).ToNot(gomega.HaveOccurred(), "controller", failedCtrl)

failedWebhook, err := webhooks.Setup(mgr)

gomega.Expect(err).ToNot(gomega.HaveOccurred(), "webhook", failedWebhook)

err = workloadjob.SetupIndexes(ctx, mgr.GetFieldIndexer())

gomega.Expect(err).NotTo(gomega.HaveOccurred())

sched := scheduler.New(

queues, cCache, mgr.GetClient(), mgr.GetEventRecorderFor(constants.AdmissionName),

scheduler.WithPodsReadyRequeuingTimestamp(requeuingTimestamp),

)

err = sched.Start(ctx)

gomega.Expect(err).NotTo(gomega.HaveOccurred())

We just register the indexers:

kueue/test/integration/scheduler/podsready/suite_test.go

Lines 96 to 97 in 975b5f2

err = workloadjob.SetupIndexes(ctx, mgr.GetFieldIndexer())

gomega.Expect(err).NotTo(gomega.HaveOccurred())

Interesting...
We should probably move that piece of code to the Workload controller in a follow up. It doesn't belong in the job reconciler, as it is completely independent from the job. Otherwise, if anyone wants to write a custom integration, they would need to implement this.

It makes sense, but we need to remember why we put this here.
Let me open an issue for this.

Can you add a TODO to remove this status update and link to #1841?

Sure. Thank you for raising it!

alculquicondor

/lgtm
/approve

k8s-ci-robot · 2024-03-14T16:08:52Z

LGTM label has been added.

Git tree hash: 255b9d366fb6ad738bc6ce44f50e054bf1017507

k8s-ci-robot · 2024-03-14T16:08:56Z

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: alculquicondor, tenzen-y

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~~ [alculquicondor,tenzen-y]

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

alculquicondor · 2024-03-14T16:09:54Z

/hold

alculquicondor · 2024-03-14T16:16:53Z

/lgtm

k8s-ci-robot · 2024-03-14T16:16:58Z

LGTM label has been added.

Git tree hash: 1d5ed08e2f7d9626e4d9f6cc1bcae6b51e5bad99

tenzen-y · 2024-03-14T16:17:01Z

/hold

… webhook Signed-off-by: Yuki Iwai <[email protected]>

tenzen-y · 2024-03-14T16:22:30Z

@alculquicondor Could you give lgtm again?

tenzen-y · 2024-03-14T16:24:05Z

/hold cancel

alculquicondor · 2024-03-14T16:24:37Z

/lgtm

k8s-ci-robot · 2024-03-14T16:24:41Z

LGTM label has been added.

Git tree hash: 0d8aaed7c14f0efb389c215977f5b98d47beb0fd

tenzen-y · 2024-03-14T16:31:22Z

/test pull-kueue-test-integration-main
due to #1829

k8s-infra-cherrypick-robot · 2024-03-14T16:44:34Z

@tenzen-y: #1838 failed to apply on top of branch "release-0.6":

Applying: WaitForPodsReady: Reset the requeueState while reconciling instead of webhook
Using index info to reconstruct a base tree...
M	charts/kueue/templates/webhook/webhook.yaml
M	pkg/controller/core/workload_controller.go
M	pkg/webhooks/workload_webhook.go
M	pkg/webhooks/workload_webhook_test.go
M	test/integration/scheduler/podsready/scheduler_test.go
M	test/integration/webhook/workload_test.go
Falling back to patching base and 3-way merge...
Auto-merging test/integration/webhook/workload_test.go
CONFLICT (content): Merge conflict in test/integration/webhook/workload_test.go
Auto-merging test/integration/scheduler/podsready/scheduler_test.go
Auto-merging pkg/webhooks/workload_webhook_test.go
Auto-merging pkg/webhooks/workload_webhook.go
Auto-merging pkg/controller/core/workload_controller.go
Auto-merging charts/kueue/templates/webhook/webhook.yaml
error: Failed to merge in the changes.
hint: Use 'git am --show-current-patch=diff' to see the failed patch
Patch failed at 0001 WaitForPodsReady: Reset the requeueState while reconciling instead of webhook
When you have resolved this problem, run "git am --continue".
If you prefer to skip this patch, run "git am --skip" instead.
To restore the original branch and stop patching, run "git am --abort".

In response to this:

/cherry-pick release-0.6

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/test-infra repository.

tenzen-y · 2024-03-14T16:47:30Z

I'm preparing the cherry-pick PR.

… webhook (kubernetes-sigs#1838) Signed-off-by: Yuki Iwai <[email protected]>

k8s-ci-robot added release-note Denotes a PR that will be considered when it comes time to generate release notes. kind/bug Categorizes issue or PR as related to a bug. cncf-cla: yes Indicates the PR's author has signed the CNCF CLA. labels Mar 13, 2024

k8s-ci-robot requested review from denkensk and mimowo March 13, 2024 23:20

k8s-ci-robot added approved Indicates a PR has been approved by an approver from all required OWNERS files. size/M Denotes a PR that changes 30-99 lines, ignoring generated files. labels Mar 13, 2024

tenzen-y force-pushed the reset-requeueState-while-reconciling branch from ffba87c to 975b5f2 Compare March 13, 2024 23:31

gabesaba reviewed Mar 14, 2024

View reviewed changes

alculquicondor reviewed Mar 14, 2024

View reviewed changes

k8s-ci-robot assigned alculquicondor Mar 14, 2024

k8s-ci-robot added the lgtm "Looks good to me", indicates that a PR is ready to be merged. label Mar 14, 2024

k8s-ci-robot added the do-not-merge/hold Indicates that a PR should not merge because someone has issued a /hold command. label Mar 14, 2024

tenzen-y force-pushed the reset-requeueState-while-reconciling branch from 975b5f2 to 105c016 Compare March 14, 2024 16:15

k8s-ci-robot removed the lgtm "Looks good to me", indicates that a PR is ready to be merged. label Mar 14, 2024

k8s-ci-robot requested a review from alculquicondor March 14, 2024 16:15

tenzen-y force-pushed the reset-requeueState-while-reconciling branch from 105c016 to 70ac2d3 Compare March 14, 2024 16:16

k8s-ci-robot added the lgtm "Looks good to me", indicates that a PR is ready to be merged. label Mar 14, 2024

tenzen-y force-pushed the reset-requeueState-while-reconciling branch from 70ac2d3 to 62bb01f Compare March 14, 2024 16:17

k8s-ci-robot removed the lgtm "Looks good to me", indicates that a PR is ready to be merged. label Mar 14, 2024

WaitForPodsReady: Reset the requeueState while reconciling instead of…

688f223

… webhook Signed-off-by: Yuki Iwai <[email protected]>

tenzen-y force-pushed the reset-requeueState-while-reconciling branch from 62bb01f to 688f223 Compare March 14, 2024 16:22

k8s-ci-robot removed the do-not-merge/hold Indicates that a PR should not merge because someone has issued a /hold command. label Mar 14, 2024

k8s-ci-robot added the lgtm "Looks good to me", indicates that a PR is ready to be merged. label Mar 14, 2024

tenzen-y mentioned this pull request Mar 14, 2024

Flaky integration test "RayCluster Job controller interacting with scheduler Should schedule jobs as they fit in their ClusterQueue" #1829

Closed

k8s-ci-robot merged commit 7ccc556 into kubernetes-sigs:main Mar 14, 2024
14 checks passed

k8s-ci-robot added this to the v0.7 milestone Mar 14, 2024

tenzen-y deleted the reset-requeueState-while-reconciling branch March 14, 2024 16:45

tenzen-y mentioned this pull request Mar 14, 2024

Automated cherry pick of #1838: WaitForPodsReady: Reset the requeueState while reconciling #1843

Merged

vsoch pushed a commit to researchapps/kueue that referenced this pull request Apr 18, 2024

WaitForPodsReady: Reset the requeueState while reconciling instead of…

a8ba152

… webhook (kubernetes-sigs#1838) Signed-off-by: Yuki Iwai <[email protected]>

tenzen-y mentioned this pull request May 10, 2024

WaitForPodsReady: After a Workload is deactivated by exceeded the backoffLimit, the reactivated Workload could be immediately re-deactivated #2174

Closed

gabesaba mentioned this pull request Jun 10, 2024

REQUEST: New membership for gabesaba kubernetes/org#5008

Closed

11 tasks

kannon92 pushed a commit to openshift-kannon92/kubernetes-sigs-kueue that referenced this pull request Nov 19, 2024

WaitForPodsReady: Reset the requeueState while reconciling instead of…

a67e77a

… webhook (kubernetes-sigs#1838) Signed-off-by: Yuki Iwai <[email protected]>

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

WaitForPodsReady: Reset the requeueState while reconciling #1838

WaitForPodsReady: Reset the requeueState while reconciling #1838

tenzen-y commented Mar 13, 2024

netlify bot commented Mar 13, 2024 •

edited

Loading

tenzen-y commented Mar 13, 2024

k8s-infra-cherrypick-robot commented Mar 13, 2024

gabesaba Mar 14, 2024

tenzen-y Mar 14, 2024 •

edited

Loading

gabesaba Mar 14, 2024

tenzen-y Mar 14, 2024

alculquicondor Mar 14, 2024

tenzen-y Mar 14, 2024

tenzen-y Mar 14, 2024

alculquicondor Mar 14, 2024

tenzen-y Mar 14, 2024 •

edited

Loading

alculquicondor Mar 14, 2024

tenzen-y Mar 14, 2024 •

edited

Loading

alculquicondor left a comment

k8s-ci-robot commented Mar 14, 2024

k8s-ci-robot commented Mar 14, 2024

alculquicondor commented Mar 14, 2024

alculquicondor commented Mar 14, 2024

k8s-ci-robot commented Mar 14, 2024

tenzen-y commented Mar 14, 2024

tenzen-y commented Mar 14, 2024

tenzen-y commented Mar 14, 2024

alculquicondor commented Mar 14, 2024

k8s-ci-robot commented Mar 14, 2024

tenzen-y commented Mar 14, 2024

k8s-infra-cherrypick-robot commented Mar 14, 2024

tenzen-y commented Mar 14, 2024

	cfg := &config.Configuration{
	WaitForPodsReady: &config.WaitForPodsReady{
	Enable: true,
	BlockAdmission: &blockAdmission,
	Timeout: &metav1.Duration{Duration: value},
	RequeuingStrategy: &config.RequeuingStrategy{
	Timestamp: ptr.To(requeuingTimestamp),
	BackoffLimitCount: requeuingBackoffLimitCount,
	},
	},
	}
	mgr.GetScheme().Default(cfg)

	err := indexer.Setup(ctx, mgr.GetFieldIndexer())
	gomega.Expect(err).NotTo(gomega.HaveOccurred())

	cCache := cache.New(mgr.GetClient(), cache.WithPodsReadyTracking(cfg.WaitForPodsReady.Enable && cfg.WaitForPodsReady.BlockAdmission != nil && *cfg.WaitForPodsReady.BlockAdmission))
	queues := queue.NewManager(
	mgr.GetClient(), cCache,
	queue.WithPodsReadyRequeuingTimestamp(requeuingTimestamp),
	)

	failedCtrl, err := core.SetupControllers(mgr, queues, cCache, cfg)
	gomega.Expect(err).ToNot(gomega.HaveOccurred(), "controller", failedCtrl)

	failedWebhook, err := webhooks.Setup(mgr)
	gomega.Expect(err).ToNot(gomega.HaveOccurred(), "webhook", failedWebhook)

	err = workloadjob.SetupIndexes(ctx, mgr.GetFieldIndexer())
	gomega.Expect(err).NotTo(gomega.HaveOccurred())

	sched := scheduler.New(
	queues, cCache, mgr.GetClient(), mgr.GetEventRecorderFor(constants.AdmissionName),
	scheduler.WithPodsReadyRequeuingTimestamp(requeuingTimestamp),
	)
	err = sched.Start(ctx)
	gomega.Expect(err).NotTo(gomega.HaveOccurred())

WaitForPodsReady: Reset the requeueState while reconciling #1838

WaitForPodsReady: Reset the requeueState while reconciling #1838

Conversation

tenzen-y commented Mar 13, 2024

What type of PR is this?

What this PR does / why we need it:

Which issue(s) this PR fixes:

Special notes for your reviewer:

Does this PR introduce a user-facing change?

netlify bot commented Mar 13, 2024 • edited Loading

✅ Deploy Preview for kubernetes-sigs-kueue canceled.

tenzen-y commented Mar 13, 2024

k8s-infra-cherrypick-robot commented Mar 13, 2024

Choose a reason for hiding this comment

tenzen-y Mar 14, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

tenzen-y Mar 14, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

tenzen-y Mar 14, 2024 • edited Loading

Choose a reason for hiding this comment

alculquicondor left a comment

Choose a reason for hiding this comment

k8s-ci-robot commented Mar 14, 2024

k8s-ci-robot commented Mar 14, 2024

alculquicondor commented Mar 14, 2024

alculquicondor commented Mar 14, 2024

k8s-ci-robot commented Mar 14, 2024

tenzen-y commented Mar 14, 2024

tenzen-y commented Mar 14, 2024

tenzen-y commented Mar 14, 2024

alculquicondor commented Mar 14, 2024

k8s-ci-robot commented Mar 14, 2024

tenzen-y commented Mar 14, 2024

k8s-infra-cherrypick-robot commented Mar 14, 2024

tenzen-y commented Mar 14, 2024

netlify bot commented Mar 13, 2024 •

edited

Loading

tenzen-y Mar 14, 2024 •

edited

Loading

tenzen-y Mar 14, 2024 •

edited

Loading

tenzen-y Mar 14, 2024 •

edited

Loading