upstream/kubernetes
1
0
Fork 0
mirror of https://github.com/kubernetes/kubernetes.git synced 2026-05-28 04:04:39 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions

KEP-3939: Job Pod Replacement Policy; promote to GA

Browse source 
Signed-off-by: Dejan Zele Pejchev <pejcev.dejan@gmail.com>
This commit is contained in:
Dejan Zele Pejchev 2025-02-13 01:49:59 +01:00
parent fc198b92c0
commit bccc9fe470
No known key found for this signature in database
GPG key ID: 8A900F09C964845E
15 changed files with 55 additions and 16 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
api/openapi-spec/swagger.json generated
View file

@ -4690,7 +4690,7 @@
"description": "Specifies the policy of handling failed pods. In particular, it allows to specify the set of actions and conditions which need to be satisfied to take the associated action. If empty, the default behaviour applies - the counter of failed pods, represented by the jobs's .status.failed field, is incremented and it is checked against the backoffLimit. This field cannot be used in combination with restartPolicy=OnFailure."
},
"podReplacementPolicy": {
"description": "podReplacementPolicy specifies when to create replacement Pods. Possible values are: - TerminatingOrFailed means that we recreate pods\n when they are terminating (has a metadata.deletionTimestamp) or failed.\n- Failed means to wait until a previously created Pod is fully terminated (has phase\n Failed or Succeeded) before creating a replacement Pod.\n\nWhen using podFailurePolicy, Failed is the the only allowed value. TerminatingOrFailed and Failed are allowed values when podFailurePolicy is not in use. This is an beta field. To use this, enable the JobPodReplacementPolicy feature toggle. This is on by default.",
"description": "podReplacementPolicy specifies when to create replacement Pods. Possible values are: - TerminatingOrFailed means that we recreate pods\n when they are terminating (has a metadata.deletionTimestamp) or failed.\n- Failed means to wait until a previously created Pod is fully terminated (has phase\n Failed or Succeeded) before creating a replacement Pod.\n\nWhen using podFailurePolicy, Failed is the the only allowed value. TerminatingOrFailed and Failed are allowed values when podFailurePolicy is not in use.",
"type": "string"
},
"selector": {

2
api/openapi-spec/v3/apis__batch__v1_openapi.json generated
View file

@ -371,7 +371,7 @@
"description": "Specifies the policy of handling failed pods. In particular, it allows to specify the set of actions and conditions which need to be satisfied to take the associated action. If empty, the default behaviour applies - the counter of failed pods, represented by the jobs's .status.failed field, is incremented and it is checked against the backoffLimit. This field cannot be used in combination with restartPolicy=OnFailure."
},
"podReplacementPolicy": {
"description": "podReplacementPolicy specifies when to create replacement Pods. Possible values are: - TerminatingOrFailed means that we recreate pods\n when they are terminating (has a metadata.deletionTimestamp) or failed.\n- Failed means to wait until a previously created Pod is fully terminated (has phase\n Failed or Succeeded) before creating a replacement Pod.\n\nWhen using podFailurePolicy, Failed is the the only allowed value. TerminatingOrFailed and Failed are allowed values when podFailurePolicy is not in use. This is an beta field. To use this, enable the JobPodReplacementPolicy feature toggle. This is on by default.",
"description": "podReplacementPolicy specifies when to create replacement Pods. Possible values are: - TerminatingOrFailed means that we recreate pods\n when they are terminating (has a metadata.deletionTimestamp) or failed.\n- Failed means to wait until a previously created Pod is fully terminated (has phase\n Failed or Succeeded) before creating a replacement Pod.\n\nWhen using podFailurePolicy, Failed is the the only allowed value. TerminatingOrFailed and Failed are allowed values when podFailurePolicy is not in use.",
"type": "string"
},
"selector": {

2
pkg/apis/batch/types.go
View file

@ -454,8 +454,6 @@ type JobSpec struct {
//
// When using podFailurePolicy, Failed is the the only allowed value.
// TerminatingOrFailed and Failed are allowed values when podFailurePolicy is not in use.
// This is an beta field. To use this, enable the JobPodReplacementPolicy feature toggle.
// This is on by default.
// +optional
PodReplacementPolicy *PodReplacementPolicy

5
pkg/apis/batch/v1/defaults_test.go
View file

@ -26,6 +26,7 @@ import (
v1 "k8s.io/api/core/v1"
metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
"k8s.io/apimachinery/pkg/runtime"
utilversion "k8s.io/apimachinery/pkg/util/version"
utilfeature "k8s.io/apiserver/pkg/util/feature"
featuregatetesting "k8s.io/component-base/featuregate/testing"
"k8s.io/kubernetes/pkg/api/legacyscheme"
@ -487,6 +488,10 @@ func TestSetDefaultJob(t *testing.T) {
for name, test := range tests {
t.Run(name, func(t *testing.T) {
if !test.enablePodReplacementPolicy {
// TODO: this will be removed in 1.37.
featuregatetesting.SetFeatureGateEmulationVersionDuringTest(t, utilfeature.DefaultFeatureGate, utilversion.MustParse("1.33"))
}
featuregatetesting.SetFeatureGateDuringTest(t, utilfeature.DefaultFeatureGate, features.JobPodReplacementPolicy, test.enablePodReplacementPolicy)
original := test.original
expected := test.expected

18
pkg/controller/job/job_controller_test.go
View file

@ -1256,6 +1256,9 @@ func TestControllerSyncJob(t *testing.T) {
} else if !tc.jobSuccessPolicy {
// TODO: this will be removed in 1.36.
featuregatetesting.SetFeatureGateEmulationVersionDuringTest(t, feature.DefaultFeatureGate, utilversion.MustParse("1.32"))
} else if !tc.jobPodReplacementPolicy {
// TODO: this will be removed in 1.37.
featuregatetesting.SetFeatureGateEmulationVersionDuringTest(t, feature.DefaultFeatureGate, utilversion.MustParse("1.33"))
}
featuregatetesting.SetFeatureGateDuringTest(t, feature.DefaultFeatureGate, features.PodIndexLabel, !tc.podIndexLabelDisabled)
featuregatetesting.SetFeatureGateDuringTest(t, feature.DefaultFeatureGate, features.JobPodReplacementPolicy, tc.jobPodReplacementPolicy)
@ -2697,6 +2700,10 @@ func TestSyncJobPastDeadline(t *testing.T) {
for name, tc := range testCases {
t.Run(name, func(t *testing.T) {
if !tc.enableJobPodReplacementPolicy {
// TODO: this will be removed in 1.37.
featuregatetesting.SetFeatureGateEmulationVersionDuringTest(t, feature.DefaultFeatureGate, utilversion.MustParse("1.33"))
}
featuregatetesting.SetFeatureGateDuringTest(t, feature.DefaultFeatureGate, features.JobManagedBy, tc.enableJobManagedBy)
featuregatetesting.SetFeatureGateDuringTest(t, feature.DefaultFeatureGate, features.JobPodReplacementPolicy, tc.enableJobPodReplacementPolicy)
// job manager setup
@ -4169,6 +4176,10 @@ func TestSyncJobWithJobPodFailurePolicy(t *testing.T) {
}
for name, tc := range testCases {
t.Run(name, func(t *testing.T) {
if !tc.enableJobPodReplacementPolicy {
// TODO: this will be removed in 1.37.
featuregatetesting.SetFeatureGateEmulationVersionDuringTest(t, feature.DefaultFeatureGate, utilversion.MustParse("1.33"))
}
featuregatetesting.SetFeatureGateDuringTest(t, feature.DefaultFeatureGate, features.JobPodReplacementPolicy, tc.enableJobPodReplacementPolicy)
featuregatetesting.SetFeatureGateDuringTest(t, feature.DefaultFeatureGate, features.JobManagedBy, tc.enableJobManagedBy)
@ -5863,6 +5874,9 @@ func TestSyncJobWithJobBackoffLimitPerIndex(t *testing.T) {
if !tc.enableJobBackoffLimitPerIndex {
// TODO: this will be removed in 1.36
featuregatetesting.SetFeatureGateEmulationVersionDuringTest(t, feature.DefaultFeatureGate, utilversion.MustParse("1.32"))
} else if !tc.enableJobPodReplacementPolicy {
// TODO: this will be removed in 1.37.
featuregatetesting.SetFeatureGateEmulationVersionDuringTest(t, feature.DefaultFeatureGate, utilversion.MustParse("1.33"))
}
featuregatetesting.SetFeatureGateDuringTest(t, feature.DefaultFeatureGate, features.JobBackoffLimitPerIndex, tc.enableJobBackoffLimitPerIndex)
featuregatetesting.SetFeatureGateDuringTest(t, feature.DefaultFeatureGate, features.JobPodReplacementPolicy, tc.enableJobPodReplacementPolicy)
@ -7365,6 +7379,10 @@ func TestJobBackoffForOnFailure(t *testing.T) {
for name, tc := range testCases {
t.Run(name, func(t *testing.T) {
if !tc.enableJobPodReplacementPolicy {
// TODO: this will be removed in 1.37.
featuregatetesting.SetFeatureGateEmulationVersionDuringTest(t, feature.DefaultFeatureGate, utilversion.MustParse("1.33"))
}
featuregatetesting.SetFeatureGateDuringTest(t, feature.DefaultFeatureGate, features.JobPodReplacementPolicy, tc.enableJobPodReplacementPolicy)
featuregatetesting.SetFeatureGateDuringTest(t, feature.DefaultFeatureGate, features.JobManagedBy, tc.enableJobManagedBy)
// job manager setup

1
pkg/features/kube_features.go
View file

@ -1400,6 +1400,7 @@ var defaultVersionedKubernetesFeatureGates = map[featuregate.Feature]featuregate
JobPodReplacementPolicy: {
{Version: version.MustParse("1.28"), Default: false, PreRelease: featuregate.Alpha},
{Version: version.MustParse("1.29"), Default: true, PreRelease: featuregate.Beta},
{Version: version.MustParse("1.34"), Default: true, PreRelease: featuregate.GA, LockToDefault: true}, // remove in 1.37
},
JobSuccessPolicy: {

2
pkg/generated/openapi/zz_generated.openapi.go generated
View file

@ -17824,7 +17824,7 @@ func schema_k8sio_api_batch_v1_JobSpec(ref common.ReferenceCallback) common.Open
},
"podReplacementPolicy": {
SchemaProps: spec.SchemaProps{
Description: "podReplacementPolicy specifies when to create replacement Pods. Possible values are: - TerminatingOrFailed means that we recreate pods\n when they are terminating (has a metadata.deletionTimestamp) or failed.\n- Failed means to wait until a previously created Pod is fully terminated (has phase\n Failed or Succeeded) before creating a replacement Pod.\n\nWhen using podFailurePolicy, Failed is the the only allowed value. TerminatingOrFailed and Failed are allowed values when podFailurePolicy is not in use. This is an beta field. To use this, enable the JobPodReplacementPolicy feature toggle. This is on by default.\n\nPossible enum values:\n - `\"Failed\"` means to wait until a previously created Pod is fully terminated (has phase Failed or Succeeded) before creating a replacement Pod.\n - `\"TerminatingOrFailed\"` means that we recreate pods when they are terminating (has a metadata.deletionTimestamp) or failed.",
Description: "podReplacementPolicy specifies when to create replacement Pods. Possible values are: - TerminatingOrFailed means that we recreate pods\n when they are terminating (has a metadata.deletionTimestamp) or failed.\n- Failed means to wait until a previously created Pod is fully terminated (has phase\n Failed or Succeeded) before creating a replacement Pod.\n\nWhen using podFailurePolicy, Failed is the the only allowed value. TerminatingOrFailed and Failed are allowed values when podFailurePolicy is not in use.\n\nPossible enum values:\n - `\"Failed\"` means to wait until a previously created Pod is fully terminated (has phase Failed or Succeeded) before creating a replacement Pod.\n - `\"TerminatingOrFailed\"` means that we recreate pods when they are terminating (has a metadata.deletionTimestamp) or failed.",
Type: []string{"string"},
Format: "",
Enum: []interface{}{"Failed", "TerminatingOrFailed"},

6
pkg/registry/batch/job/strategy_test.go
View file

@ -516,6 +516,9 @@ func TestJobStrategy_PrepareForUpdate(t *testing.T) {
if !tc.enableJobBackoffLimitPerIndex || !tc.enableJobSuccessPolicy {
// TODO: this will be removed in 1.36
featuregatetesting.SetFeatureGateEmulationVersionDuringTest(t, utilfeature.DefaultFeatureGate, utilversion.MustParse("1.32"))
} else if !tc.enableJobPodReplacementPolicy {
// TODO: this will be removed in 1.37.
featuregatetesting.SetFeatureGateEmulationVersionDuringTest(t, utilfeature.DefaultFeatureGate, utilversion.MustParse("1.33"))
}
featuregatetesting.SetFeatureGateDuringTest(t, utilfeature.DefaultFeatureGate, features.JobBackoffLimitPerIndex, tc.enableJobBackoffLimitPerIndex)
featuregatetesting.SetFeatureGateDuringTest(t, utilfeature.DefaultFeatureGate, features.JobPodReplacementPolicy, tc.enableJobPodReplacementPolicy)
@ -3541,6 +3544,9 @@ func TestStatusStrategy_ValidateUpdate(t *testing.T) {
if !tc.enableJobSuccessPolicy {
// TODO: this will be removed in 1.36.
featuregatetesting.SetFeatureGateEmulationVersionDuringTest(t, utilfeature.DefaultFeatureGate, utilversion.MustParse("1.32"))
} else if !tc.enableJobPodReplacementPolicy {
// TODO: this will be removed in 1.37.
featuregatetesting.SetFeatureGateEmulationVersionDuringTest(t, utilfeature.DefaultFeatureGate, utilversion.MustParse("1.33"))
}
featuregatetesting.SetFeatureGateDuringTest(t, utilfeature.DefaultFeatureGate, features.JobManagedBy, tc.enableJobManagedBy)
featuregatetesting.SetFeatureGateDuringTest(t, utilfeature.DefaultFeatureGate, features.JobSuccessPolicy, tc.enableJobSuccessPolicy)

2
staging/src/k8s.io/api/batch/v1/generated.proto
View file

@ -330,8 +330,6 @@ message JobSpec {
//
// When using podFailurePolicy, Failed is the the only allowed value.
// TerminatingOrFailed and Failed are allowed values when podFailurePolicy is not in use.
// This is an beta field. To use this, enable the JobPodReplacementPolicy feature toggle.
// This is on by default.
// +optional
optional string podReplacementPolicy = 14;

2
staging/src/k8s.io/api/batch/v1/types.go
View file

@ -456,8 +456,6 @@ type JobSpec struct {
//
// When using podFailurePolicy, Failed is the the only allowed value.
// TerminatingOrFailed and Failed are allowed values when podFailurePolicy is not in use.
// This is an beta field. To use this, enable the JobPodReplacementPolicy feature toggle.
// This is on by default.
// +optional
PodReplacementPolicy *PodReplacementPolicy `json:"podReplacementPolicy,omitempty" protobuf:"bytes,14,opt,name=podReplacementPolicy,casttype=podReplacementPolicy"`

2
staging/src/k8s.io/api/batch/v1/types_swagger_doc_generated.go generated
View file

@ -126,7 +126,7 @@ var map_JobSpec = map[string]string{
"ttlSecondsAfterFinished": "ttlSecondsAfterFinished limits the lifetime of a Job that has finished execution (either Complete or Failed). If this field is set, ttlSecondsAfterFinished after the Job finishes, it is eligible to be automatically deleted. When the Job is being deleted, its lifecycle guarantees (e.g. finalizers) will be honored. If this field is unset, the Job won't be automatically deleted. If this field is set to zero, the Job becomes eligible to be deleted immediately after it finishes.",
"completionMode": "completionMode specifies how Pod completions are tracked. It can be `NonIndexed` (default) or `Indexed`.\n\n`NonIndexed` means that the Job is considered complete when there have been .spec.completions successfully completed Pods. Each Pod completion is homologous to each other.\n\n`Indexed` means that the Pods of a Job get an associated completion index from 0 to (.spec.completions - 1), available in the annotation batch.kubernetes.io/job-completion-index. The Job is considered complete when there is one successfully completed Pod for each index. When value is `Indexed`, .spec.completions must be specified and `.spec.parallelism` must be less than or equal to 10^5. In addition, The Pod name takes the form `$(job-name)-$(index)-$(random-string)`, the Pod hostname takes the form `$(job-name)-$(index)`.\n\nMore completion modes can be added in the future. If the Job controller observes a mode that it doesn't recognize, which is possible during upgrades due to version skew, the controller skips updates for the Job.",
"suspend": "suspend specifies whether the Job controller should create Pods or not. If a Job is created with suspend set to true, no Pods are created by the Job controller. If a Job is suspended after creation (i.e. the flag goes from false to true), the Job controller will delete all active Pods associated with this Job. Users must design their workload to gracefully handle this. Suspending a Job will reset the StartTime field of the Job, effectively resetting the ActiveDeadlineSeconds timer too. Defaults to false.",
"podReplacementPolicy": "podReplacementPolicy specifies when to create replacement Pods. Possible values are: - TerminatingOrFailed means that we recreate pods\n when they are terminating (has a metadata.deletionTimestamp) or failed.\n- Failed means to wait until a previously created Pod is fully terminated (has phase\n Failed or Succeeded) before creating a replacement Pod.\n\nWhen using podFailurePolicy, Failed is the the only allowed value. TerminatingOrFailed and Failed are allowed values when podFailurePolicy is not in use. This is an beta field. To use this, enable the JobPodReplacementPolicy feature toggle. This is on by default.",
"podReplacementPolicy": "podReplacementPolicy specifies when to create replacement Pods. Possible values are: - TerminatingOrFailed means that we recreate pods\n when they are terminating (has a metadata.deletionTimestamp) or failed.\n- Failed means to wait until a previously created Pod is fully terminated (has phase\n Failed or Succeeded) before creating a replacement Pod.\n\nWhen using podFailurePolicy, Failed is the the only allowed value. TerminatingOrFailed and Failed are allowed values when podFailurePolicy is not in use.",
"managedBy": "ManagedBy field indicates the controller that manages a Job. The k8s Job controller reconciles jobs which don't have this field at all or the field value is the reserved string `kubernetes.io/job-controller`, but skips reconciling Jobs with a custom value for this field. The value must be a valid domain-prefixed path (e.g. acme.io/foo) - all characters before the first \"/\" must be a valid subdomain as defined by RFC 1123. All characters trailing the first \"/\" must be valid HTTP Path characters as defined by RFC 3986. The value cannot exceed 63 characters. This field is immutable.\n\nThis field is beta-level. The job controller accepts setting the field when the feature gate JobManagedBy is enabled (enabled by default).",
}

4
test/compatibility_lifecycle/reference/versioned_feature_list.yaml
View file

@ -645,6 +645,10 @@
lockToDefault: false
preRelease: Beta
version: "1.29"
- default: true
lockToDefault: true
preRelease: GA
version: "1.34"
- name: JobSuccessPolicy
versionedSpecs:
- default: false

9
test/e2e/apps/job.go
View file

@ -511,8 +511,7 @@ done`}
framework.ExpectNoError(err, "failed to get latest job object")
gomega.Expect(job.Status.Active).Should(gomega.Equal(int32(0)))
gomega.Expect(job.Status.Ready).Should(gomega.Equal(ptr.To[int32](0)))
// TODO (https://github.com/kubernetes/enhancements/issues/3939): Restore the assert on .status.terminating when PodReplacementPolicy goes GA.
// gomega.Expect(job.Status.Terminating).Should(gomega.Equal(ptr.To[int32](0)))
gomega.Expect(job.Status.Terminating).Should(gomega.Equal(ptr.To[int32](0)))
gomega.Expect(job.Status.Failed).Should(gomega.Equal(int32(0)))
})
@ -553,8 +552,7 @@ done`}
gomega.Expect(job.Status.CompletedIndexes).Should(gomega.Equal("0"))
gomega.Expect(job.Status.Active).Should(gomega.Equal(int32(0)))
gomega.Expect(job.Status.Ready).Should(gomega.Equal(ptr.To[int32](0)))
// TODO (https://github.com/kubernetes/enhancements/issues/3939): Restore the assert on .status.terminating when PodReplacementPolicy goes GA.
// gomega.Expect(job.Status.Terminating).Should(gomega.Equal(ptr.To[int32](0)))
gomega.Expect(job.Status.Terminating).Should(gomega.Equal(ptr.To[int32](0)))
})
/*
@ -594,8 +592,7 @@ done`}
gomega.Expect(job.Status.CompletedIndexes).Should(gomega.Equal("0"))
gomega.Expect(job.Status.Active).Should(gomega.Equal(int32(0)))
gomega.Expect(job.Status.Ready).Should(gomega.Equal(ptr.To[int32](0)))
// TODO (https://github.com/kubernetes/enhancements/issues/3939): Restore the assert on .status.terminating when PodReplacementPolicy goes GA.
// gomega.Expect(job.Status.Terminating).Should(gomega.Equal(ptr.To[int32](0)))
gomega.Expect(job.Status.Terminating).Should(gomega.Equal(ptr.To[int32](0)))
})
/*

9
test/integration/job/job_test.go
View file

@ -1588,6 +1588,9 @@ func TestDelayTerminalPhaseCondition(t *testing.T) {
if !test.enableJobSuccessPolicy {
// TODO: this will be removed in 1.36.
featuregatetesting.SetFeatureGateEmulationVersionDuringTest(t, feature.DefaultFeatureGate, utilversion.MustParse("1.32"))
} else if !test.enableJobPodReplacementPolicy {
// TODO: this will be removed in 1.37.
featuregatetesting.SetFeatureGateEmulationVersionDuringTest(t, feature.DefaultFeatureGate, utilversion.MustParse("1.33"))
}
featuregatetesting.SetFeatureGateDuringTest(t, feature.DefaultFeatureGate, features.JobPodReplacementPolicy, test.enableJobPodReplacementPolicy)
featuregatetesting.SetFeatureGateDuringTest(t, feature.DefaultFeatureGate, features.JobManagedBy, test.enableJobManagedBy)
@ -3196,6 +3199,10 @@ func TestJobPodReplacementPolicy(t *testing.T) {
for name, tc := range cases {
tc := tc
t.Run(name, func(t *testing.T) {
if !tc.podReplacementPolicyEnabled {
// TODO: this will be removed in 1.37.
featuregatetesting.SetFeatureGateEmulationVersionDuringTest(t, feature.DefaultFeatureGate, utilversion.MustParse("1.33"))
}
featuregatetesting.SetFeatureGateDuringTest(t, feature.DefaultFeatureGate, features.JobPodReplacementPolicy, tc.podReplacementPolicyEnabled)
ctx, cancel := startJobControllerAndWaitForCaches(t, restConfig)
@ -3262,6 +3269,8 @@ func TestJobPodReplacementPolicy(t *testing.T) {
// Enabling will then match the original failed case.
func TestJobPodReplacementPolicyFeatureToggling(t *testing.T) {
t.Cleanup(setDurationDuringTest(&jobcontroller.DefaultJobPodFailureBackOff, fastPodFailureBackoff))
// TODO: this will be removed in 1.37.
featuregatetesting.SetFeatureGateEmulationVersionDuringTest(t, feature.DefaultFeatureGate, utilversion.MustParse("1.33"))
const podCount int32 = 2
jobSpec := batchv1.JobSpec{
Parallelism: ptr.To(podCount),

5
test/integration/podgc/podgc_test.go
View file

@ -25,6 +25,7 @@ import (
v1 "k8s.io/api/core/v1"
apierrors "k8s.io/apimachinery/pkg/api/errors"
metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
utilversion "k8s.io/apimachinery/pkg/util/version"
"k8s.io/apimachinery/pkg/util/wait"
utilfeature "k8s.io/apiserver/pkg/util/feature"
"k8s.io/client-go/informers"
@ -160,6 +161,10 @@ func TestTerminatingOnOutOfServiceNode(t *testing.T) {
for name, test := range tests {
t.Run(name, func(t *testing.T) {
if !test.enableJobPodReplacementPolicy {
// TODO: this will be removed in 1.37.
featuregatetesting.SetFeatureGateEmulationVersionDuringTest(t, utilfeature.DefaultFeatureGate, utilversion.MustParse("1.33"))
}
featuregatetesting.SetFeatureGateDuringTest(t, utilfeature.DefaultFeatureGate, features.JobPodReplacementPolicy, test.enableJobPodReplacementPolicy)
testCtx := setup(t, "podgc-out-of-service")
cs := testCtx.ClientSet