rebase: update kubernetes to v1.20.0

updated kubernetes packages to latest
release.

Signed-off-by: Madhu Rajanna <madhupr007@gmail.com>
This commit is contained in:
Madhu Rajanna
2020-12-17 17:58:29 +05:30
committed by mergify[bot]
parent 4abe128bd8
commit 83559144b1
1624 changed files with 247222 additions and 160270 deletions

View File

@ -3,4 +3,3 @@
reviewers:
- justinsb
- freehan
- david-mcmahon

View File

@ -32,6 +32,7 @@ go_test(
"//staging/src/k8s.io/apimachinery/pkg/util/validation/field:go_default_library",
"//staging/src/k8s.io/apiserver/pkg/util/feature:go_default_library",
"//staging/src/k8s.io/component-base/featuregate/testing:go_default_library",
"//vendor/github.com/google/go-cmp/cmp:go_default_library",
"//vendor/github.com/stretchr/testify/assert:go_default_library",
],
)

View File

@ -50,28 +50,60 @@ func FindPort(pod *v1.Pod, svcPort *v1.ServicePort) (int, error) {
return 0, fmt.Errorf("no suitable port for manifest: %s", pod.UID)
}
// ContainerType signifies container type
type ContainerType int
const (
// Containers is for normal containers
Containers ContainerType = 1 << iota
// InitContainers is for init containers
InitContainers
// EphemeralContainers is for ephemeral containers
EphemeralContainers
)
// AllContainers specifies that all containers be visited
const AllContainers ContainerType = (InitContainers | Containers | EphemeralContainers)
// AllFeatureEnabledContainers returns a ContainerType mask which includes all container
// types except for the ones guarded by feature gate.
func AllFeatureEnabledContainers() ContainerType {
containerType := AllContainers
if !utilfeature.DefaultFeatureGate.Enabled(features.EphemeralContainers) {
containerType &= ^EphemeralContainers
}
return containerType
}
// ContainerVisitor is called with each container spec, and returns true
// if visiting should continue.
type ContainerVisitor func(container *v1.Container) (shouldContinue bool)
type ContainerVisitor func(container *v1.Container, containerType ContainerType) (shouldContinue bool)
// VisitContainers invokes the visitor function with a pointer to the container
// spec of every container in the given pod spec. If visitor returns false,
// Visitor is called with each object name, and returns true if visiting should continue
type Visitor func(name string) (shouldContinue bool)
// VisitContainers invokes the visitor function with a pointer to every container
// spec in the given pod spec with type set in mask. If visitor returns false,
// visiting is short-circuited. VisitContainers returns true if visiting completes,
// false if visiting was short-circuited.
func VisitContainers(podSpec *v1.PodSpec, visitor ContainerVisitor) bool {
for i := range podSpec.InitContainers {
if !visitor(&podSpec.InitContainers[i]) {
return false
func VisitContainers(podSpec *v1.PodSpec, mask ContainerType, visitor ContainerVisitor) bool {
if mask&InitContainers != 0 {
for i := range podSpec.InitContainers {
if !visitor(&podSpec.InitContainers[i], InitContainers) {
return false
}
}
}
for i := range podSpec.Containers {
if !visitor(&podSpec.Containers[i]) {
return false
if mask&Containers != 0 {
for i := range podSpec.Containers {
if !visitor(&podSpec.Containers[i], Containers) {
return false
}
}
}
if utilfeature.DefaultFeatureGate.Enabled(features.EphemeralContainers) {
if mask&EphemeralContainers != 0 {
for i := range podSpec.EphemeralContainers {
if !visitor((*v1.Container)(&podSpec.EphemeralContainers[i].EphemeralContainerCommon)) {
if !visitor((*v1.Container)(&podSpec.EphemeralContainers[i].EphemeralContainerCommon), EphemeralContainers) {
return false
}
}
@ -79,9 +111,6 @@ func VisitContainers(podSpec *v1.PodSpec, visitor ContainerVisitor) bool {
return true
}
// Visitor is called with each object name, and returns true if visiting should continue
type Visitor func(name string) (shouldContinue bool)
// VisitPodSecretNames invokes the visitor function with the name of every secret
// referenced by the pod spec. If visitor returns false, visiting is short-circuited.
// Transitive references (e.g. pod -> pvc -> pv -> secret) are not visited.
@ -92,7 +121,7 @@ func VisitPodSecretNames(pod *v1.Pod, visitor Visitor) bool {
return false
}
}
VisitContainers(&pod.Spec, func(c *v1.Container) bool {
VisitContainers(&pod.Spec, AllContainers, func(c *v1.Container, containerType ContainerType) bool {
return visitContainerSecretNames(c, visitor)
})
var source *v1.VolumeSource
@ -176,7 +205,7 @@ func visitContainerSecretNames(container *v1.Container, visitor Visitor) bool {
// Transitive references (e.g. pod -> pvc -> pv -> secret) are not visited.
// Returns true if visiting completed, false if visiting was short-circuited.
func VisitPodConfigmapNames(pod *v1.Pod, visitor Visitor) bool {
VisitContainers(&pod.Spec, func(c *v1.Container) bool {
VisitContainers(&pod.Spec, AllContainers, func(c *v1.Container, containerType ContainerType) bool {
return visitContainerConfigmapNames(c, visitor)
})
var source *v1.VolumeSource
@ -323,14 +352,3 @@ func UpdatePodCondition(status *v1.PodStatus, condition *v1.PodCondition) bool {
// Return true if one of the fields have changed.
return !isEqual
}
// GetPodPriority returns priority of the given pod.
func GetPodPriority(pod *v1.Pod) int32 {
if pod.Spec.Priority != nil {
return *pod.Spec.Priority
}
// When priority of a running pod is nil, it means it was created at a time
// that there was no global default priority class and the priority class
// name of the pod was empty. So, we resolve to the static default priority.
return 0
}

View File

@ -10,13 +10,11 @@ reviewers:
- sttts
- saad-ali
- ncdc
- tallclair
- dims
- errordeveloper
- mml
- m1093782566
- mbohlool
- david-mcmahon
- kevin-wangzefeng
- jianhuiz
labels:

View File

@ -1,57 +0,0 @@
package(default_visibility = ["//visibility:public"])
load(
"@io_bazel_rules_go//go:def.bzl",
"go_library",
"go_test",
)
go_library(
name = "go_default_library",
srcs = ["validation.go"],
importpath = "k8s.io/kubernetes/pkg/apis/apps/validation",
deps = [
"//pkg/apis/apps:go_default_library",
"//pkg/apis/core:go_default_library",
"//pkg/apis/core/validation:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/api/equality:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/api/validation:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/apis/meta/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/apis/meta/v1/validation:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/labels:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/intstr:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/validation:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/validation/field:go_default_library",
],
)
go_test(
name = "go_default_test",
srcs = ["validation_test.go"],
embed = [":go_default_library"],
deps = [
"//pkg/apis/apps:go_default_library",
"//pkg/apis/core:go_default_library",
"//pkg/features:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/apis/meta/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/runtime:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/intstr:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/validation/field:go_default_library",
"//staging/src/k8s.io/apiserver/pkg/util/feature:go_default_library",
"//staging/src/k8s.io/component-base/featuregate/testing:go_default_library",
"//vendor/github.com/davecgh/go-spew/spew:go_default_library",
],
)
filegroup(
name = "package-srcs",
srcs = glob(["**"]),
tags = ["automanaged"],
visibility = ["//visibility:private"],
)
filegroup(
name = "all-srcs",
srcs = [":package-srcs"],
tags = ["automanaged"],
)

View File

@ -1,687 +0,0 @@
/*
Copyright 2016 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 validation
import (
"fmt"
"strconv"
apiequality "k8s.io/apimachinery/pkg/api/equality"
apimachineryvalidation "k8s.io/apimachinery/pkg/api/validation"
metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
unversionedvalidation "k8s.io/apimachinery/pkg/apis/meta/v1/validation"
"k8s.io/apimachinery/pkg/labels"
"k8s.io/apimachinery/pkg/util/intstr"
"k8s.io/apimachinery/pkg/util/validation"
"k8s.io/apimachinery/pkg/util/validation/field"
"k8s.io/kubernetes/pkg/apis/apps"
api "k8s.io/kubernetes/pkg/apis/core"
apivalidation "k8s.io/kubernetes/pkg/apis/core/validation"
)
// ValidateStatefulSetName can be used to check whether the given StatefulSet name is valid.
// Prefix indicates this name will be used as part of generation, in which case
// trailing dashes are allowed.
func ValidateStatefulSetName(name string, prefix bool) []string {
// TODO: Validate that there's name for the suffix inserted by the pods.
// Currently this is just "-index". In the future we may allow a user
// specified list of suffixes and we need to validate the longest one.
return apimachineryvalidation.NameIsDNSSubdomain(name, prefix)
}
// ValidatePodTemplateSpecForStatefulSet validates the given template and ensures that it is in accordance with the desired selector.
func ValidatePodTemplateSpecForStatefulSet(template *api.PodTemplateSpec, selector labels.Selector, fldPath *field.Path) field.ErrorList {
allErrs := field.ErrorList{}
if template == nil {
allErrs = append(allErrs, field.Required(fldPath, ""))
} else {
if !selector.Empty() {
// Verify that the StatefulSet selector matches the labels in template.
labels := labels.Set(template.Labels)
if !selector.Matches(labels) {
allErrs = append(allErrs, field.Invalid(fldPath.Child("metadata", "labels"), template.Labels, "`selector` does not match template `labels`"))
}
}
// TODO: Add validation for PodSpec, currently this will check volumes, which we know will
// fail. We should really check that the union of the given volumes and volumeClaims match
// volume mounts in the containers.
// allErrs = append(allErrs, apivalidation.ValidatePodTemplateSpec(template, fldPath)...)
allErrs = append(allErrs, unversionedvalidation.ValidateLabels(template.Labels, fldPath.Child("labels"))...)
allErrs = append(allErrs, apivalidation.ValidateAnnotations(template.Annotations, fldPath.Child("annotations"))...)
allErrs = append(allErrs, apivalidation.ValidatePodSpecificAnnotations(template.Annotations, &template.Spec, fldPath.Child("annotations"))...)
}
return allErrs
}
// ValidateStatefulSetSpec tests if required fields in the StatefulSet spec are set.
func ValidateStatefulSetSpec(spec *apps.StatefulSetSpec, fldPath *field.Path) field.ErrorList {
allErrs := field.ErrorList{}
switch spec.PodManagementPolicy {
case "":
allErrs = append(allErrs, field.Required(fldPath.Child("podManagementPolicy"), ""))
case apps.OrderedReadyPodManagement, apps.ParallelPodManagement:
default:
allErrs = append(allErrs, field.Invalid(fldPath.Child("podManagementPolicy"), spec.PodManagementPolicy, fmt.Sprintf("must be '%s' or '%s'", apps.OrderedReadyPodManagement, apps.ParallelPodManagement)))
}
switch spec.UpdateStrategy.Type {
case "":
allErrs = append(allErrs, field.Required(fldPath.Child("updateStrategy"), ""))
case apps.OnDeleteStatefulSetStrategyType:
if spec.UpdateStrategy.RollingUpdate != nil {
allErrs = append(
allErrs,
field.Invalid(
fldPath.Child("updateStrategy").Child("rollingUpdate"),
spec.UpdateStrategy.RollingUpdate,
fmt.Sprintf("only allowed for updateStrategy '%s'", apps.RollingUpdateStatefulSetStrategyType)))
}
case apps.RollingUpdateStatefulSetStrategyType:
if spec.UpdateStrategy.RollingUpdate != nil {
allErrs = append(allErrs,
apivalidation.ValidateNonnegativeField(
int64(spec.UpdateStrategy.RollingUpdate.Partition),
fldPath.Child("updateStrategy").Child("rollingUpdate").Child("partition"))...)
}
default:
allErrs = append(allErrs,
field.Invalid(fldPath.Child("updateStrategy"), spec.UpdateStrategy,
fmt.Sprintf("must be '%s' or '%s'",
apps.RollingUpdateStatefulSetStrategyType,
apps.OnDeleteStatefulSetStrategyType)))
}
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(spec.Replicas), fldPath.Child("replicas"))...)
if spec.Selector == nil {
allErrs = append(allErrs, field.Required(fldPath.Child("selector"), ""))
} else {
allErrs = append(allErrs, unversionedvalidation.ValidateLabelSelector(spec.Selector, fldPath.Child("selector"))...)
if len(spec.Selector.MatchLabels)+len(spec.Selector.MatchExpressions) == 0 {
allErrs = append(allErrs, field.Invalid(fldPath.Child("selector"), spec.Selector, "empty selector is invalid for statefulset"))
}
}
selector, err := metav1.LabelSelectorAsSelector(spec.Selector)
if err != nil {
allErrs = append(allErrs, field.Invalid(fldPath.Child("selector"), spec.Selector, ""))
} else {
allErrs = append(allErrs, ValidatePodTemplateSpecForStatefulSet(&spec.Template, selector, fldPath.Child("template"))...)
}
if spec.Template.Spec.RestartPolicy != api.RestartPolicyAlways {
allErrs = append(allErrs, field.NotSupported(fldPath.Child("template", "spec", "restartPolicy"), spec.Template.Spec.RestartPolicy, []string{string(api.RestartPolicyAlways)}))
}
if spec.Template.Spec.ActiveDeadlineSeconds != nil {
allErrs = append(allErrs, field.Forbidden(fldPath.Child("template", "spec", "activeDeadlineSeconds"), "activeDeadlineSeconds in StatefulSet is not Supported"))
}
return allErrs
}
// ValidateStatefulSet validates a StatefulSet.
func ValidateStatefulSet(statefulSet *apps.StatefulSet) field.ErrorList {
allErrs := apivalidation.ValidateObjectMeta(&statefulSet.ObjectMeta, true, ValidateStatefulSetName, field.NewPath("metadata"))
allErrs = append(allErrs, ValidateStatefulSetSpec(&statefulSet.Spec, field.NewPath("spec"))...)
return allErrs
}
// ValidateStatefulSetUpdate tests if required fields in the StatefulSet are set.
func ValidateStatefulSetUpdate(statefulSet, oldStatefulSet *apps.StatefulSet) field.ErrorList {
allErrs := apivalidation.ValidateObjectMetaUpdate(&statefulSet.ObjectMeta, &oldStatefulSet.ObjectMeta, field.NewPath("metadata"))
restoreReplicas := statefulSet.Spec.Replicas
statefulSet.Spec.Replicas = oldStatefulSet.Spec.Replicas
restoreTemplate := statefulSet.Spec.Template
statefulSet.Spec.Template = oldStatefulSet.Spec.Template
restoreStrategy := statefulSet.Spec.UpdateStrategy
statefulSet.Spec.UpdateStrategy = oldStatefulSet.Spec.UpdateStrategy
if !apiequality.Semantic.DeepEqual(statefulSet.Spec, oldStatefulSet.Spec) {
allErrs = append(allErrs, field.Forbidden(field.NewPath("spec"), "updates to statefulset spec for fields other than 'replicas', 'template', and 'updateStrategy' are forbidden"))
}
statefulSet.Spec.Replicas = restoreReplicas
statefulSet.Spec.Template = restoreTemplate
statefulSet.Spec.UpdateStrategy = restoreStrategy
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(statefulSet.Spec.Replicas), field.NewPath("spec", "replicas"))...)
return allErrs
}
// ValidateStatefulSetStatus validates a StatefulSetStatus.
func ValidateStatefulSetStatus(status *apps.StatefulSetStatus, fieldPath *field.Path) field.ErrorList {
allErrs := field.ErrorList{}
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(status.Replicas), fieldPath.Child("replicas"))...)
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(status.ReadyReplicas), fieldPath.Child("readyReplicas"))...)
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(status.CurrentReplicas), fieldPath.Child("currentReplicas"))...)
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(status.UpdatedReplicas), fieldPath.Child("updatedReplicas"))...)
if status.ObservedGeneration != nil {
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(*status.ObservedGeneration), fieldPath.Child("observedGeneration"))...)
}
if status.CollisionCount != nil {
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(*status.CollisionCount), fieldPath.Child("collisionCount"))...)
}
msg := "cannot be greater than status.replicas"
if status.ReadyReplicas > status.Replicas {
allErrs = append(allErrs, field.Invalid(fieldPath.Child("readyReplicas"), status.ReadyReplicas, msg))
}
if status.CurrentReplicas > status.Replicas {
allErrs = append(allErrs, field.Invalid(fieldPath.Child("currentReplicas"), status.CurrentReplicas, msg))
}
if status.UpdatedReplicas > status.Replicas {
allErrs = append(allErrs, field.Invalid(fieldPath.Child("updatedReplicas"), status.UpdatedReplicas, msg))
}
return allErrs
}
// ValidateStatefulSetStatusUpdate tests if required fields in the StatefulSet are set.
func ValidateStatefulSetStatusUpdate(statefulSet, oldStatefulSet *apps.StatefulSet) field.ErrorList {
allErrs := field.ErrorList{}
allErrs = append(allErrs, ValidateStatefulSetStatus(&statefulSet.Status, field.NewPath("status"))...)
allErrs = append(allErrs, apivalidation.ValidateObjectMetaUpdate(&statefulSet.ObjectMeta, &oldStatefulSet.ObjectMeta, field.NewPath("metadata"))...)
// TODO: Validate status.
if apivalidation.IsDecremented(statefulSet.Status.CollisionCount, oldStatefulSet.Status.CollisionCount) {
value := int32(0)
if statefulSet.Status.CollisionCount != nil {
value = *statefulSet.Status.CollisionCount
}
allErrs = append(allErrs, field.Invalid(field.NewPath("status").Child("collisionCount"), value, "cannot be decremented"))
}
return allErrs
}
// ValidateControllerRevisionName can be used to check whether the given ControllerRevision name is valid.
// Prefix indicates this name will be used as part of generation, in which case
// trailing dashes are allowed.
var ValidateControllerRevisionName = apimachineryvalidation.NameIsDNSSubdomain
// ValidateControllerRevision collects errors for the fields of state and returns those errors as an ErrorList. If the
// returned list is empty, state is valid. Validation is performed to ensure that state is a valid ObjectMeta, its name
// is valid, and that it doesn't exceed the MaxControllerRevisionSize.
func ValidateControllerRevision(revision *apps.ControllerRevision) field.ErrorList {
errs := field.ErrorList{}
errs = append(errs, apivalidation.ValidateObjectMeta(&revision.ObjectMeta, true, ValidateControllerRevisionName, field.NewPath("metadata"))...)
if revision.Data == nil {
errs = append(errs, field.Required(field.NewPath("data"), "data is mandatory"))
}
errs = append(errs, apivalidation.ValidateNonnegativeField(revision.Revision, field.NewPath("revision"))...)
return errs
}
// ValidateControllerRevisionUpdate collects errors pertaining to the mutation of an ControllerRevision Object. If the
// returned ErrorList is empty the update operation is valid. Any mutation to the ControllerRevision's Data or Revision
// is considered to be invalid.
func ValidateControllerRevisionUpdate(newHistory, oldHistory *apps.ControllerRevision) field.ErrorList {
errs := field.ErrorList{}
errs = append(errs, apivalidation.ValidateObjectMetaUpdate(&newHistory.ObjectMeta, &oldHistory.ObjectMeta, field.NewPath("metadata"))...)
errs = append(errs, ValidateControllerRevision(newHistory)...)
errs = append(errs, apivalidation.ValidateImmutableField(newHistory.Data, oldHistory.Data, field.NewPath("data"))...)
return errs
}
// ValidateDaemonSet tests if required fields in the DaemonSet are set.
func ValidateDaemonSet(ds *apps.DaemonSet) field.ErrorList {
allErrs := apivalidation.ValidateObjectMeta(&ds.ObjectMeta, true, ValidateDaemonSetName, field.NewPath("metadata"))
allErrs = append(allErrs, ValidateDaemonSetSpec(&ds.Spec, field.NewPath("spec"))...)
return allErrs
}
// ValidateDaemonSetUpdate tests if required fields in the DaemonSet are set.
func ValidateDaemonSetUpdate(ds, oldDS *apps.DaemonSet) field.ErrorList {
allErrs := apivalidation.ValidateObjectMetaUpdate(&ds.ObjectMeta, &oldDS.ObjectMeta, field.NewPath("metadata"))
allErrs = append(allErrs, ValidateDaemonSetSpecUpdate(&ds.Spec, &oldDS.Spec, field.NewPath("spec"))...)
allErrs = append(allErrs, ValidateDaemonSetSpec(&ds.Spec, field.NewPath("spec"))...)
return allErrs
}
// ValidateDaemonSetSpecUpdate tests if an update to a DaemonSetSpec is valid.
func ValidateDaemonSetSpecUpdate(newSpec, oldSpec *apps.DaemonSetSpec, fldPath *field.Path) field.ErrorList {
allErrs := field.ErrorList{}
// TemplateGeneration shouldn't be decremented
if newSpec.TemplateGeneration < oldSpec.TemplateGeneration {
allErrs = append(allErrs, field.Invalid(fldPath.Child("templateGeneration"), newSpec.TemplateGeneration, "must not be decremented"))
}
// TemplateGeneration should be increased when and only when template is changed
templateUpdated := !apiequality.Semantic.DeepEqual(newSpec.Template, oldSpec.Template)
if newSpec.TemplateGeneration == oldSpec.TemplateGeneration && templateUpdated {
allErrs = append(allErrs, field.Invalid(fldPath.Child("templateGeneration"), newSpec.TemplateGeneration, "must be incremented upon template update"))
} else if newSpec.TemplateGeneration > oldSpec.TemplateGeneration && !templateUpdated {
allErrs = append(allErrs, field.Invalid(fldPath.Child("templateGeneration"), newSpec.TemplateGeneration, "must not be incremented without template update"))
}
return allErrs
}
// validateDaemonSetStatus validates a DaemonSetStatus
func validateDaemonSetStatus(status *apps.DaemonSetStatus, fldPath *field.Path) field.ErrorList {
allErrs := field.ErrorList{}
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(status.CurrentNumberScheduled), fldPath.Child("currentNumberScheduled"))...)
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(status.NumberMisscheduled), fldPath.Child("numberMisscheduled"))...)
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(status.DesiredNumberScheduled), fldPath.Child("desiredNumberScheduled"))...)
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(status.NumberReady), fldPath.Child("numberReady"))...)
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(status.ObservedGeneration, fldPath.Child("observedGeneration"))...)
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(status.UpdatedNumberScheduled), fldPath.Child("updatedNumberScheduled"))...)
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(status.NumberAvailable), fldPath.Child("numberAvailable"))...)
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(status.NumberUnavailable), fldPath.Child("numberUnavailable"))...)
if status.CollisionCount != nil {
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(*status.CollisionCount), fldPath.Child("collisionCount"))...)
}
return allErrs
}
// ValidateDaemonSetStatusUpdate tests if required fields in the DaemonSet Status section
func ValidateDaemonSetStatusUpdate(ds, oldDS *apps.DaemonSet) field.ErrorList {
allErrs := apivalidation.ValidateObjectMetaUpdate(&ds.ObjectMeta, &oldDS.ObjectMeta, field.NewPath("metadata"))
allErrs = append(allErrs, validateDaemonSetStatus(&ds.Status, field.NewPath("status"))...)
if apivalidation.IsDecremented(ds.Status.CollisionCount, oldDS.Status.CollisionCount) {
value := int32(0)
if ds.Status.CollisionCount != nil {
value = *ds.Status.CollisionCount
}
allErrs = append(allErrs, field.Invalid(field.NewPath("status").Child("collisionCount"), value, "cannot be decremented"))
}
return allErrs
}
// ValidateDaemonSetSpec tests if required fields in the DaemonSetSpec are set.
func ValidateDaemonSetSpec(spec *apps.DaemonSetSpec, fldPath *field.Path) field.ErrorList {
allErrs := field.ErrorList{}
allErrs = append(allErrs, unversionedvalidation.ValidateLabelSelector(spec.Selector, fldPath.Child("selector"))...)
selector, err := metav1.LabelSelectorAsSelector(spec.Selector)
if err == nil && !selector.Matches(labels.Set(spec.Template.Labels)) {
allErrs = append(allErrs, field.Invalid(fldPath.Child("template", "metadata", "labels"), spec.Template.Labels, "`selector` does not match template `labels`"))
}
if spec.Selector != nil && len(spec.Selector.MatchLabels)+len(spec.Selector.MatchExpressions) == 0 {
allErrs = append(allErrs, field.Invalid(fldPath.Child("selector"), spec.Selector, "empty selector is invalid for daemonset"))
}
allErrs = append(allErrs, apivalidation.ValidatePodTemplateSpec(&spec.Template, fldPath.Child("template"))...)
// Daemons typically run on more than one node, so mark Read-Write persistent disks as invalid.
allErrs = append(allErrs, apivalidation.ValidateReadOnlyPersistentDisks(spec.Template.Spec.Volumes, fldPath.Child("template", "spec", "volumes"))...)
// RestartPolicy has already been first-order validated as per ValidatePodTemplateSpec().
if spec.Template.Spec.RestartPolicy != api.RestartPolicyAlways {
allErrs = append(allErrs, field.NotSupported(fldPath.Child("template", "spec", "restartPolicy"), spec.Template.Spec.RestartPolicy, []string{string(api.RestartPolicyAlways)}))
}
if spec.Template.Spec.ActiveDeadlineSeconds != nil {
allErrs = append(allErrs, field.Forbidden(fldPath.Child("template", "spec", "activeDeadlineSeconds"), "activeDeadlineSeconds in DaemonSet is not Supported"))
}
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(spec.MinReadySeconds), fldPath.Child("minReadySeconds"))...)
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(spec.TemplateGeneration), fldPath.Child("templateGeneration"))...)
allErrs = append(allErrs, ValidateDaemonSetUpdateStrategy(&spec.UpdateStrategy, fldPath.Child("updateStrategy"))...)
if spec.RevisionHistoryLimit != nil {
// zero is a valid RevisionHistoryLimit
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(*spec.RevisionHistoryLimit), fldPath.Child("revisionHistoryLimit"))...)
}
return allErrs
}
// ValidateRollingUpdateDaemonSet validates a given RollingUpdateDaemonSet.
func ValidateRollingUpdateDaemonSet(rollingUpdate *apps.RollingUpdateDaemonSet, fldPath *field.Path) field.ErrorList {
allErrs := field.ErrorList{}
allErrs = append(allErrs, ValidatePositiveIntOrPercent(rollingUpdate.MaxUnavailable, fldPath.Child("maxUnavailable"))...)
if getIntOrPercentValue(rollingUpdate.MaxUnavailable) == 0 {
// MaxUnavailable cannot be 0.
allErrs = append(allErrs, field.Invalid(fldPath.Child("maxUnavailable"), rollingUpdate.MaxUnavailable, "cannot be 0"))
}
// Validate that MaxUnavailable is not more than 100%.
allErrs = append(allErrs, IsNotMoreThan100Percent(rollingUpdate.MaxUnavailable, fldPath.Child("maxUnavailable"))...)
return allErrs
}
// ValidateDaemonSetUpdateStrategy validates a given DaemonSetUpdateStrategy.
func ValidateDaemonSetUpdateStrategy(strategy *apps.DaemonSetUpdateStrategy, fldPath *field.Path) field.ErrorList {
allErrs := field.ErrorList{}
switch strategy.Type {
case apps.OnDeleteDaemonSetStrategyType:
case apps.RollingUpdateDaemonSetStrategyType:
// Make sure RollingUpdate field isn't nil.
if strategy.RollingUpdate == nil {
allErrs = append(allErrs, field.Required(fldPath.Child("rollingUpdate"), ""))
return allErrs
}
allErrs = append(allErrs, ValidateRollingUpdateDaemonSet(strategy.RollingUpdate, fldPath.Child("rollingUpdate"))...)
default:
validValues := []string{string(apps.RollingUpdateDaemonSetStrategyType), string(apps.OnDeleteDaemonSetStrategyType)}
allErrs = append(allErrs, field.NotSupported(fldPath, strategy, validValues))
}
return allErrs
}
// ValidateDaemonSetName can be used to check whether the given daemon set name is valid.
// Prefix indicates this name will be used as part of generation, in which case
// trailing dashes are allowed.
var ValidateDaemonSetName = apimachineryvalidation.NameIsDNSSubdomain
// ValidateDeploymentName validates that the given name can be used as a deployment name.
var ValidateDeploymentName = apimachineryvalidation.NameIsDNSSubdomain
// ValidatePositiveIntOrPercent tests if a given value is a valid int or
// percentage.
func ValidatePositiveIntOrPercent(intOrPercent intstr.IntOrString, fldPath *field.Path) field.ErrorList {
allErrs := field.ErrorList{}
switch intOrPercent.Type {
case intstr.String:
for _, msg := range validation.IsValidPercent(intOrPercent.StrVal) {
allErrs = append(allErrs, field.Invalid(fldPath, intOrPercent, msg))
}
case intstr.Int:
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(intOrPercent.IntValue()), fldPath)...)
default:
allErrs = append(allErrs, field.Invalid(fldPath, intOrPercent, "must be an integer or percentage (e.g '5%%')"))
}
return allErrs
}
func getPercentValue(intOrStringValue intstr.IntOrString) (int, bool) {
if intOrStringValue.Type != intstr.String {
return 0, false
}
if len(validation.IsValidPercent(intOrStringValue.StrVal)) != 0 {
return 0, false
}
value, _ := strconv.Atoi(intOrStringValue.StrVal[:len(intOrStringValue.StrVal)-1])
return value, true
}
func getIntOrPercentValue(intOrStringValue intstr.IntOrString) int {
value, isPercent := getPercentValue(intOrStringValue)
if isPercent {
return value
}
return intOrStringValue.IntValue()
}
// IsNotMoreThan100Percent tests is a value can be represented as a percentage
// and if this value is not more than 100%.
func IsNotMoreThan100Percent(intOrStringValue intstr.IntOrString, fldPath *field.Path) field.ErrorList {
allErrs := field.ErrorList{}
value, isPercent := getPercentValue(intOrStringValue)
if !isPercent || value <= 100 {
return nil
}
allErrs = append(allErrs, field.Invalid(fldPath, intOrStringValue, "must not be greater than 100%"))
return allErrs
}
// ValidateRollingUpdateDeployment validates a given RollingUpdateDeployment.
func ValidateRollingUpdateDeployment(rollingUpdate *apps.RollingUpdateDeployment, fldPath *field.Path) field.ErrorList {
allErrs := field.ErrorList{}
allErrs = append(allErrs, ValidatePositiveIntOrPercent(rollingUpdate.MaxUnavailable, fldPath.Child("maxUnavailable"))...)
allErrs = append(allErrs, ValidatePositiveIntOrPercent(rollingUpdate.MaxSurge, fldPath.Child("maxSurge"))...)
if getIntOrPercentValue(rollingUpdate.MaxUnavailable) == 0 && getIntOrPercentValue(rollingUpdate.MaxSurge) == 0 {
// Both MaxSurge and MaxUnavailable cannot be zero.
allErrs = append(allErrs, field.Invalid(fldPath.Child("maxUnavailable"), rollingUpdate.MaxUnavailable, "may not be 0 when `maxSurge` is 0"))
}
// Validate that MaxUnavailable is not more than 100%.
allErrs = append(allErrs, IsNotMoreThan100Percent(rollingUpdate.MaxUnavailable, fldPath.Child("maxUnavailable"))...)
return allErrs
}
// ValidateDeploymentStrategy validates given DeploymentStrategy.
func ValidateDeploymentStrategy(strategy *apps.DeploymentStrategy, fldPath *field.Path) field.ErrorList {
allErrs := field.ErrorList{}
switch strategy.Type {
case apps.RecreateDeploymentStrategyType:
if strategy.RollingUpdate != nil {
allErrs = append(allErrs, field.Forbidden(fldPath.Child("rollingUpdate"), "may not be specified when strategy `type` is '"+string(apps.RecreateDeploymentStrategyType+"'")))
}
case apps.RollingUpdateDeploymentStrategyType:
// This should never happen since it's set and checked in defaults.go
if strategy.RollingUpdate == nil {
allErrs = append(allErrs, field.Required(fldPath.Child("rollingUpdate"), "this should be defaulted and never be nil"))
} else {
allErrs = append(allErrs, ValidateRollingUpdateDeployment(strategy.RollingUpdate, fldPath.Child("rollingUpdate"))...)
}
default:
validValues := []string{string(apps.RecreateDeploymentStrategyType), string(apps.RollingUpdateDeploymentStrategyType)}
allErrs = append(allErrs, field.NotSupported(fldPath, strategy, validValues))
}
return allErrs
}
// ValidateRollback validates given RollbackConfig.
func ValidateRollback(rollback *apps.RollbackConfig, fldPath *field.Path) field.ErrorList {
allErrs := field.ErrorList{}
v := rollback.Revision
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(v), fldPath.Child("version"))...)
return allErrs
}
// ValidateDeploymentSpec validates given deployment spec.
func ValidateDeploymentSpec(spec *apps.DeploymentSpec, fldPath *field.Path) field.ErrorList {
allErrs := field.ErrorList{}
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(spec.Replicas), fldPath.Child("replicas"))...)
if spec.Selector == nil {
allErrs = append(allErrs, field.Required(fldPath.Child("selector"), ""))
} else {
allErrs = append(allErrs, unversionedvalidation.ValidateLabelSelector(spec.Selector, fldPath.Child("selector"))...)
if len(spec.Selector.MatchLabels)+len(spec.Selector.MatchExpressions) == 0 {
allErrs = append(allErrs, field.Invalid(fldPath.Child("selector"), spec.Selector, "empty selector is invalid for deployment"))
}
}
selector, err := metav1.LabelSelectorAsSelector(spec.Selector)
if err != nil {
allErrs = append(allErrs, field.Invalid(fldPath.Child("selector"), spec.Selector, "invalid label selector"))
} else {
allErrs = append(allErrs, ValidatePodTemplateSpecForReplicaSet(&spec.Template, selector, spec.Replicas, fldPath.Child("template"))...)
}
allErrs = append(allErrs, ValidateDeploymentStrategy(&spec.Strategy, fldPath.Child("strategy"))...)
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(spec.MinReadySeconds), fldPath.Child("minReadySeconds"))...)
if spec.RevisionHistoryLimit != nil {
// zero is a valid RevisionHistoryLimit
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(*spec.RevisionHistoryLimit), fldPath.Child("revisionHistoryLimit"))...)
}
if spec.RollbackTo != nil {
allErrs = append(allErrs, ValidateRollback(spec.RollbackTo, fldPath.Child("rollback"))...)
}
if spec.ProgressDeadlineSeconds != nil {
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(*spec.ProgressDeadlineSeconds), fldPath.Child("progressDeadlineSeconds"))...)
if *spec.ProgressDeadlineSeconds <= spec.MinReadySeconds {
allErrs = append(allErrs, field.Invalid(fldPath.Child("progressDeadlineSeconds"), spec.ProgressDeadlineSeconds, "must be greater than minReadySeconds"))
}
}
return allErrs
}
// ValidateDeploymentStatus validates given deployment status.
func ValidateDeploymentStatus(status *apps.DeploymentStatus, fldPath *field.Path) field.ErrorList {
allErrs := field.ErrorList{}
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(status.ObservedGeneration, fldPath.Child("observedGeneration"))...)
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(status.Replicas), fldPath.Child("replicas"))...)
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(status.UpdatedReplicas), fldPath.Child("updatedReplicas"))...)
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(status.ReadyReplicas), fldPath.Child("readyReplicas"))...)
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(status.AvailableReplicas), fldPath.Child("availableReplicas"))...)
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(status.UnavailableReplicas), fldPath.Child("unavailableReplicas"))...)
if status.CollisionCount != nil {
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(*status.CollisionCount), fldPath.Child("collisionCount"))...)
}
msg := "cannot be greater than status.replicas"
if status.UpdatedReplicas > status.Replicas {
allErrs = append(allErrs, field.Invalid(fldPath.Child("updatedReplicas"), status.UpdatedReplicas, msg))
}
if status.ReadyReplicas > status.Replicas {
allErrs = append(allErrs, field.Invalid(fldPath.Child("readyReplicas"), status.ReadyReplicas, msg))
}
if status.AvailableReplicas > status.Replicas {
allErrs = append(allErrs, field.Invalid(fldPath.Child("availableReplicas"), status.AvailableReplicas, msg))
}
if status.AvailableReplicas > status.ReadyReplicas {
allErrs = append(allErrs, field.Invalid(fldPath.Child("availableReplicas"), status.AvailableReplicas, "cannot be greater than readyReplicas"))
}
return allErrs
}
// ValidateDeploymentUpdate tests if an update to a Deployment is valid.
func ValidateDeploymentUpdate(update, old *apps.Deployment) field.ErrorList {
allErrs := apivalidation.ValidateObjectMetaUpdate(&update.ObjectMeta, &old.ObjectMeta, field.NewPath("metadata"))
allErrs = append(allErrs, ValidateDeploymentSpec(&update.Spec, field.NewPath("spec"))...)
return allErrs
}
// ValidateDeploymentStatusUpdate tests if a an update to a Deployment status
// is valid.
func ValidateDeploymentStatusUpdate(update, old *apps.Deployment) field.ErrorList {
allErrs := apivalidation.ValidateObjectMetaUpdate(&update.ObjectMeta, &old.ObjectMeta, field.NewPath("metadata"))
fldPath := field.NewPath("status")
allErrs = append(allErrs, ValidateDeploymentStatus(&update.Status, fldPath)...)
if apivalidation.IsDecremented(update.Status.CollisionCount, old.Status.CollisionCount) {
value := int32(0)
if update.Status.CollisionCount != nil {
value = *update.Status.CollisionCount
}
allErrs = append(allErrs, field.Invalid(fldPath.Child("collisionCount"), value, "cannot be decremented"))
}
return allErrs
}
// ValidateDeployment validates a given Deployment.
func ValidateDeployment(obj *apps.Deployment) field.ErrorList {
allErrs := apivalidation.ValidateObjectMeta(&obj.ObjectMeta, true, ValidateDeploymentName, field.NewPath("metadata"))
allErrs = append(allErrs, ValidateDeploymentSpec(&obj.Spec, field.NewPath("spec"))...)
return allErrs
}
// ValidateDeploymentRollback validates a given DeploymentRollback.
func ValidateDeploymentRollback(obj *apps.DeploymentRollback) field.ErrorList {
allErrs := apivalidation.ValidateAnnotations(obj.UpdatedAnnotations, field.NewPath("updatedAnnotations"))
if len(obj.Name) == 0 {
allErrs = append(allErrs, field.Required(field.NewPath("name"), "name is required"))
}
allErrs = append(allErrs, ValidateRollback(&obj.RollbackTo, field.NewPath("rollback"))...)
return allErrs
}
// ValidateReplicaSetName can be used to check whether the given ReplicaSet
// name is valid.
// Prefix indicates this name will be used as part of generation, in which case
// trailing dashes are allowed.
var ValidateReplicaSetName = apimachineryvalidation.NameIsDNSSubdomain
// ValidateReplicaSet tests if required fields in the ReplicaSet are set.
func ValidateReplicaSet(rs *apps.ReplicaSet) field.ErrorList {
allErrs := apivalidation.ValidateObjectMeta(&rs.ObjectMeta, true, ValidateReplicaSetName, field.NewPath("metadata"))
allErrs = append(allErrs, ValidateReplicaSetSpec(&rs.Spec, field.NewPath("spec"))...)
return allErrs
}
// ValidateReplicaSetUpdate tests if required fields in the ReplicaSet are set.
func ValidateReplicaSetUpdate(rs, oldRs *apps.ReplicaSet) field.ErrorList {
allErrs := field.ErrorList{}
allErrs = append(allErrs, apivalidation.ValidateObjectMetaUpdate(&rs.ObjectMeta, &oldRs.ObjectMeta, field.NewPath("metadata"))...)
allErrs = append(allErrs, ValidateReplicaSetSpec(&rs.Spec, field.NewPath("spec"))...)
return allErrs
}
// ValidateReplicaSetStatusUpdate tests if required fields in the ReplicaSet are set.
func ValidateReplicaSetStatusUpdate(rs, oldRs *apps.ReplicaSet) field.ErrorList {
allErrs := field.ErrorList{}
allErrs = append(allErrs, apivalidation.ValidateObjectMetaUpdate(&rs.ObjectMeta, &oldRs.ObjectMeta, field.NewPath("metadata"))...)
allErrs = append(allErrs, ValidateReplicaSetStatus(rs.Status, field.NewPath("status"))...)
return allErrs
}
// ValidateReplicaSetStatus validates a given ReplicaSetStatus.
func ValidateReplicaSetStatus(status apps.ReplicaSetStatus, fldPath *field.Path) field.ErrorList {
allErrs := field.ErrorList{}
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(status.Replicas), fldPath.Child("replicas"))...)
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(status.FullyLabeledReplicas), fldPath.Child("fullyLabeledReplicas"))...)
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(status.ReadyReplicas), fldPath.Child("readyReplicas"))...)
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(status.AvailableReplicas), fldPath.Child("availableReplicas"))...)
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(status.ObservedGeneration), fldPath.Child("observedGeneration"))...)
msg := "cannot be greater than status.replicas"
if status.FullyLabeledReplicas > status.Replicas {
allErrs = append(allErrs, field.Invalid(fldPath.Child("fullyLabeledReplicas"), status.FullyLabeledReplicas, msg))
}
if status.ReadyReplicas > status.Replicas {
allErrs = append(allErrs, field.Invalid(fldPath.Child("readyReplicas"), status.ReadyReplicas, msg))
}
if status.AvailableReplicas > status.Replicas {
allErrs = append(allErrs, field.Invalid(fldPath.Child("availableReplicas"), status.AvailableReplicas, msg))
}
if status.AvailableReplicas > status.ReadyReplicas {
allErrs = append(allErrs, field.Invalid(fldPath.Child("availableReplicas"), status.AvailableReplicas, "cannot be greater than readyReplicas"))
}
return allErrs
}
// ValidateReplicaSetSpec tests if required fields in the ReplicaSet spec are set.
func ValidateReplicaSetSpec(spec *apps.ReplicaSetSpec, fldPath *field.Path) field.ErrorList {
allErrs := field.ErrorList{}
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(spec.Replicas), fldPath.Child("replicas"))...)
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(spec.MinReadySeconds), fldPath.Child("minReadySeconds"))...)
if spec.Selector == nil {
allErrs = append(allErrs, field.Required(fldPath.Child("selector"), ""))
} else {
allErrs = append(allErrs, unversionedvalidation.ValidateLabelSelector(spec.Selector, fldPath.Child("selector"))...)
if len(spec.Selector.MatchLabels)+len(spec.Selector.MatchExpressions) == 0 {
allErrs = append(allErrs, field.Invalid(fldPath.Child("selector"), spec.Selector, "empty selector is invalid for deployment"))
}
}
selector, err := metav1.LabelSelectorAsSelector(spec.Selector)
if err != nil {
allErrs = append(allErrs, field.Invalid(fldPath.Child("selector"), spec.Selector, "invalid label selector"))
} else {
allErrs = append(allErrs, ValidatePodTemplateSpecForReplicaSet(&spec.Template, selector, spec.Replicas, fldPath.Child("template"))...)
}
return allErrs
}
// ValidatePodTemplateSpecForReplicaSet validates the given template and ensures that it is in accordance with the desired selector and replicas.
func ValidatePodTemplateSpecForReplicaSet(template *api.PodTemplateSpec, selector labels.Selector, replicas int32, fldPath *field.Path) field.ErrorList {
allErrs := field.ErrorList{}
if template == nil {
allErrs = append(allErrs, field.Required(fldPath, ""))
} else {
if !selector.Empty() {
// Verify that the ReplicaSet selector matches the labels in template.
labels := labels.Set(template.Labels)
if !selector.Matches(labels) {
allErrs = append(allErrs, field.Invalid(fldPath.Child("metadata", "labels"), template.Labels, "`selector` does not match template `labels`"))
}
}
allErrs = append(allErrs, apivalidation.ValidatePodTemplateSpec(template, fldPath)...)
if replicas > 1 {
allErrs = append(allErrs, apivalidation.ValidateReadOnlyPersistentDisks(template.Spec.Volumes, fldPath.Child("spec", "volumes"))...)
}
// RestartPolicy has already been first-order validated as per ValidatePodTemplateSpec().
if template.Spec.RestartPolicy != api.RestartPolicyAlways {
allErrs = append(allErrs, field.NotSupported(fldPath.Child("spec", "restartPolicy"), template.Spec.RestartPolicy, []string{string(api.RestartPolicyAlways)}))
}
if template.Spec.ActiveDeadlineSeconds != nil {
allErrs = append(allErrs, field.Forbidden(fldPath.Child("spec", "activeDeadlineSeconds"), "activeDeadlineSeconds in ReplicaSet is not Supported"))
}
}
return allErrs
}

View File

@ -16,5 +16,4 @@ reviewers:
- madhusudancs
- mml
- mbohlool
- david-mcmahon
- jianhuiz

View File

@ -204,6 +204,12 @@ const (
// (for example length of queue in cloud messaging service, or
// QPS from loadbalancer running outside of cluster).
ExternalMetricSourceType MetricSourceType = "External"
// ContainerResourceMetricSourceType is a resource metric known to Kubernetes, as
// specified in requests and limits, describing a single container in each pod in the current
// scale target (e.g. CPU or memory). Such metrics are built in to
// Kubernetes, and have special scaling options on top of those available
// to normal per-pod metrics (the "pods" source).
ContainerResourceMetricSourceType MetricSourceType = "ContainerResource"
)
// MetricSpec specifies how to scale based on a single metric
@ -229,6 +235,13 @@ type MetricSpec struct {
// to normal per-pod metrics using the "pods" source.
// +optional
Resource *ResourceMetricSource
// ContainerResource refers to a resource metric (such as those specified in
// requests and limits) known to Kubernetes describing a single container in each pod of the
// current scale target (e.g. CPU or memory). Such metrics are built in to
// Kubernetes, and have special scaling options on top of those available
// to normal per-pod metrics using the "pods" source.
// +optional
ContainerResource *ContainerResourceMetricSource
// External refers to a global metric that is not associated
// with any Kubernetes object. It allows autoscaling based on information
// coming from components running outside of cluster
@ -271,6 +284,22 @@ type ResourceMetricSource struct {
Target MetricTarget
}
// ContainerResourceMetricSource indicates how to scale on a resource metric known to
// Kubernetes, as specified in the requests and limits, describing a single container in
// each of the pods of the current scale target(e.g. CPU or memory). The values will be
// averaged together before being compared to the target. Such metrics are built into
// Kubernetes, and have special scaling options on top of those available to
// normal per-pod metrics using the "pods" source. Only one "target" type
// should be set.
type ContainerResourceMetricSource struct {
// name is the name of the of the resource
Name api.ResourceName
// container is the name of the container in the pods of the scaling target.
Container string
// target specifies the target value for the given metric
Target MetricTarget
}
// ExternalMetricSource indicates how to scale on a metric not associated with
// any Kubernetes object (for example length of queue in cloud
// messaging service, or QPS from loadbalancer running outside of cluster).
@ -420,6 +449,13 @@ type MetricStatus struct {
// to normal per-pod metrics using the "pods" source.
// +optional
Resource *ResourceMetricStatus
// ContainerResource refers to a resource metric (such as those specified in
// requests and limits) known to Kubernetes describing a single container in each pod in the
// current scale target (e.g. CPU or memory). Such metrics are built in to
// Kubernetes, and have special scaling options on top of those available
// to normal per-pod metrics using the "pods" source.
// +optional
ContainerResource *ContainerResourceMetricStatus
// External refers to a global metric that is not associated
// with any Kubernetes object. It allows autoscaling based on information
// coming from components running outside of cluster
@ -456,6 +492,18 @@ type ResourceMetricStatus struct {
Current MetricValueStatus
}
// ContainerResourceMetricStatus indicates the current value of a resource metric known to
// Kubernetes, as specified in requests and limits, describing each pod in the
// current scale target (e.g. CPU or memory). Such metrics are built in to
// Kubernetes, and have special scaling options on top of those available to
// normal per-pod metrics using the "pods" source.
type ContainerResourceMetricStatus struct {
// Name is the name of the resource in question.
Name api.ResourceName
Container string
Current MetricValueStatus
}
// ExternalMetricStatus indicates the current value of a global metric
// not associated with any Kubernetes object.
type ExternalMetricStatus struct {

View File

@ -25,6 +25,40 @@ import (
runtime "k8s.io/apimachinery/pkg/runtime"
)
// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
func (in *ContainerResourceMetricSource) DeepCopyInto(out *ContainerResourceMetricSource) {
*out = *in
in.Target.DeepCopyInto(&out.Target)
return
}
// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new ContainerResourceMetricSource.
func (in *ContainerResourceMetricSource) DeepCopy() *ContainerResourceMetricSource {
if in == nil {
return nil
}
out := new(ContainerResourceMetricSource)
in.DeepCopyInto(out)
return out
}
// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
func (in *ContainerResourceMetricStatus) DeepCopyInto(out *ContainerResourceMetricStatus) {
*out = *in
in.Current.DeepCopyInto(&out.Current)
return
}
// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new ContainerResourceMetricStatus.
func (in *ContainerResourceMetricStatus) DeepCopy() *ContainerResourceMetricStatus {
if in == nil {
return nil
}
out := new(ContainerResourceMetricStatus)
in.DeepCopyInto(out)
return out
}
// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
func (in *CrossVersionObjectReference) DeepCopyInto(out *CrossVersionObjectReference) {
*out = *in
@ -340,6 +374,11 @@ func (in *MetricSpec) DeepCopyInto(out *MetricSpec) {
*out = new(ResourceMetricSource)
(*in).DeepCopyInto(*out)
}
if in.ContainerResource != nil {
in, out := &in.ContainerResource, &out.ContainerResource
*out = new(ContainerResourceMetricSource)
(*in).DeepCopyInto(*out)
}
if in.External != nil {
in, out := &in.External, &out.External
*out = new(ExternalMetricSource)
@ -376,6 +415,11 @@ func (in *MetricStatus) DeepCopyInto(out *MetricStatus) {
*out = new(ResourceMetricStatus)
(*in).DeepCopyInto(*out)
}
if in.ContainerResource != nil {
in, out := &in.ContainerResource, &out.ContainerResource
*out = new(ContainerResourceMetricStatus)
(*in).DeepCopyInto(*out)
}
if in.External != nil {
in, out := &in.External, &out.External
*out = new(ExternalMetricStatus)

View File

@ -16,7 +16,6 @@ reviewers:
- errordeveloper
- mml
- mbohlool
- david-mcmahon
- jianhuiz
labels:
- sig/apps

View File

@ -155,6 +155,7 @@ type JobSpec struct {
type JobStatus struct {
// The latest available observations of an object's current state.
// When a job fails, one of the conditions will have type == "Failed".
// +optional
Conditions []JobCondition
@ -167,6 +168,7 @@ type JobStatus struct {
// Represents time when the job was completed. It is not guaranteed to
// be set in happens-before order across separate operations.
// It is represented in RFC3339 form and is in UTC.
// The completion time is only set when the job finishes successfully.
// +optional
CompletionTime *metav1.Time

View File

@ -6,7 +6,6 @@ approvers:
- smarterclayton
- thockin
- liggitt
# - bgrant0607 # manual escalations only
reviewers:
- thockin
- lavalamp
@ -28,7 +27,6 @@ reviewers:
- sttts
- dchen1107
- saad-ali
- zmerlynn
- luxas
- janetkuo
- justinsb
@ -36,7 +34,6 @@ reviewers:
- ncdc
- tallclair
- yifan-gu
- eparis
- mwielgus
- soltysh
- piosz

View File

@ -39,17 +39,20 @@ const (
// SeccompPodAnnotationKey represents the key of a seccomp profile applied
// to all containers of a pod.
// Deprecated: set a pod security context `seccompProfile` field.
SeccompPodAnnotationKey string = "seccomp.security.alpha.kubernetes.io/pod"
// SeccompContainerAnnotationKeyPrefix represents the key of a seccomp profile applied
// to one container of a pod.
// Deprecated: set a container security context `seccompProfile` field.
SeccompContainerAnnotationKeyPrefix string = "container.seccomp.security.alpha.kubernetes.io/"
// SeccompProfileRuntimeDefault represents the default seccomp profile used by container runtime.
// Deprecated: set a pod or container security context `seccompProfile` of type "RuntimeDefault" instead.
SeccompProfileRuntimeDefault string = "runtime/default"
// DeprecatedSeccompProfileDockerDefault represents the default seccomp profile used by docker.
// This is now deprecated and should be replaced by SeccompProfileRuntimeDefault.
// Deprecated: set a pod or container security context `seccompProfile` of type "RuntimeDefault" instead.
DeprecatedSeccompProfileDockerDefault string = "docker/default"
// PreferAvoidPodsAnnotationKey represents the key of preferAvoidPods data (json serialized)
@ -61,10 +64,6 @@ const (
// This annotation can be attached to node.
ObjectTTLAnnotationKey string = "node.alpha.kubernetes.io/ttl"
// BootstrapCheckpointAnnotationKey represents a Resource (Pod) that should be checkpointed by
// the kubelet prior to running
BootstrapCheckpointAnnotationKey string = "node.kubernetes.io/bootstrap-checkpoint"
// NonConvertibleAnnotationPrefix annotation key prefix used to identify non-convertible json paths.
NonConvertibleAnnotationPrefix = "non-convertible.kubernetes.io"

View File

@ -267,7 +267,10 @@ func IsIntegerResourceName(str string) bool {
// IsServiceIPSet aims to check if the service's ClusterIP is set or not
// the objective is not to perform validation here
func IsServiceIPSet(service *core.Service) bool {
return service.Spec.ClusterIP != core.ClusterIPNone && service.Spec.ClusterIP != ""
// This function assumes that the service is semantically validated
// it does not test if the IP is valid, just makes sure that it is set.
return len(service.Spec.ClusterIP) > 0 &&
service.Spec.ClusterIP != core.ClusterIPNone
}
var standardFinalizers = sets.NewString(
@ -494,3 +497,38 @@ func PersistentVolumeClaimHasClass(claim *core.PersistentVolumeClaim) bool {
return false
}
func toResourceNames(resources core.ResourceList) []core.ResourceName {
result := []core.ResourceName{}
for resourceName := range resources {
result = append(result, resourceName)
}
return result
}
func toSet(resourceNames []core.ResourceName) sets.String {
result := sets.NewString()
for _, resourceName := range resourceNames {
result.Insert(string(resourceName))
}
return result
}
// toContainerResourcesSet returns a set of resources names in container resource requirements
func toContainerResourcesSet(ctr *core.Container) sets.String {
resourceNames := toResourceNames(ctr.Resources.Requests)
resourceNames = append(resourceNames, toResourceNames(ctr.Resources.Limits)...)
return toSet(resourceNames)
}
// ToPodResourcesSet returns a set of resource names in all containers in a pod.
func ToPodResourcesSet(podSpec *core.PodSpec) sets.String {
result := sets.NewString()
for i := range podSpec.InitContainers {
result = result.Union(toContainerResourcesSet(&podSpec.InitContainers[i]))
}
for i := range podSpec.Containers {
result = result.Union(toContainerResourcesSet(&podSpec.Containers[i]))
}
return result
}

View File

@ -8,4 +8,3 @@ reviewers:
- liggitt
- nikhiljindal
- dims
- david-mcmahon

View File

@ -34,23 +34,23 @@ type ContainerVisitorWithPath func(container *api.Container, path *field.Path) b
// of every container in the given pod spec and the field.Path to that container.
// If visitor returns false, visiting is short-circuited. VisitContainersWithPath returns true if visiting completes,
// false if visiting was short-circuited.
func VisitContainersWithPath(podSpec *api.PodSpec, visitor ContainerVisitorWithPath) bool {
path := field.NewPath("spec", "initContainers")
func VisitContainersWithPath(podSpec *api.PodSpec, specPath *field.Path, visitor ContainerVisitorWithPath) bool {
fldPath := specPath.Child("initContainers")
for i := range podSpec.InitContainers {
if !visitor(&podSpec.InitContainers[i], path.Index(i)) {
if !visitor(&podSpec.InitContainers[i], fldPath.Index(i)) {
return false
}
}
path = field.NewPath("spec", "containers")
fldPath = specPath.Child("containers")
for i := range podSpec.Containers {
if !visitor(&podSpec.Containers[i], path.Index(i)) {
if !visitor(&podSpec.Containers[i], fldPath.Index(i)) {
return false
}
}
if utilfeature.DefaultFeatureGate.Enabled(features.EphemeralContainers) {
path = field.NewPath("spec", "ephemeralContainers")
fldPath = specPath.Child("ephemeralContainers")
for i := range podSpec.EphemeralContainers {
if !visitor((*api.Container)(&podSpec.EphemeralContainers[i].EphemeralContainerCommon), path.Index(i)) {
if !visitor((*api.Container)(&podSpec.EphemeralContainers[i].EphemeralContainerCommon), fldPath.Index(i)) {
return false
}
}

View File

@ -26,40 +26,33 @@ func (rn ResourceName) String() string {
// CPU returns the CPU limit if specified.
func (rl *ResourceList) CPU() *resource.Quantity {
if val, ok := (*rl)[ResourceCPU]; ok {
return &val
}
return &resource.Quantity{Format: resource.DecimalSI}
return rl.Name(ResourceCPU, resource.DecimalSI)
}
// Memory returns the Memory limit if specified.
func (rl *ResourceList) Memory() *resource.Quantity {
if val, ok := (*rl)[ResourceMemory]; ok {
return &val
}
return &resource.Quantity{Format: resource.BinarySI}
return rl.Name(ResourceMemory, resource.BinarySI)
}
// Storage returns the Storage limit if specified.
func (rl *ResourceList) Storage() *resource.Quantity {
if val, ok := (*rl)[ResourceStorage]; ok {
return &val
}
return &resource.Quantity{Format: resource.BinarySI}
return rl.Name(ResourceStorage, resource.BinarySI)
}
// Pods returns the list of pods
func (rl *ResourceList) Pods() *resource.Quantity {
if val, ok := (*rl)[ResourcePods]; ok {
return &val
}
return &resource.Quantity{}
return rl.Name(ResourcePods, resource.DecimalSI)
}
// StorageEphemeral returns the list of ephemeral storage volumes, if any
func (rl *ResourceList) StorageEphemeral() *resource.Quantity {
if val, ok := (*rl)[ResourceEphemeralStorage]; ok {
return rl.Name(ResourceEphemeralStorage, resource.BinarySI)
}
// Name returns the resource with name if specified, otherwise it returns a nil quantity with default format.
func (rl *ResourceList) Name(name ResourceName, defaultFormat resource.Format) *resource.Quantity {
if val, ok := (*rl)[name]; ok {
return &val
}
return &resource.Quantity{}
return &resource.Quantity{Format: defaultFormat}
}

View File

@ -154,9 +154,36 @@ type VolumeSource struct {
// StorageOS represents a StorageOS volume that is attached to the kubelet's host machine and mounted into the pod
// +optional
StorageOS *StorageOSVolumeSource
// CSI (Container Storage Interface) represents storage that is handled by an external CSI driver (Alpha feature).
// CSI (Container Storage Interface) represents ephemeral storage that is handled by certain external CSI drivers (Beta feature).
// +optional
CSI *CSIVolumeSource
// Ephemeral represents a volume that is handled by a cluster storage driver (Alpha feature).
// The volume's lifecycle is tied to the pod that defines it - it will be created before the pod starts,
// and deleted when the pod is removed.
//
// Use this if:
// a) the volume is only needed while the pod runs,
// b) features of normal volumes like restoring from snapshot or capacity
// tracking are needed,
// c) the storage driver is specified through a storage class, and
// d) the storage driver supports dynamic volume provisioning through
// a PersistentVolumeClaim (see EphemeralVolumeSource for more
// information on the connection between this volume type
// and PersistentVolumeClaim).
//
// Use PersistentVolumeClaim or one of the vendor-specific
// APIs for volumes that persist for longer than the lifecycle
// of an individual pod.
//
// Use CSI for light-weight local ephemeral volumes if the CSI driver is meant to
// be used that way - see the documentation of the driver for
// more information.
//
// A pod can use both types of ephemeral volumes and
// persistent volumes at the same time.
//
// +optional
Ephemeral *EphemeralVolumeSource
}
// PersistentVolumeSource is similar to VolumeSource but meant for the administrator who creates PVs.
@ -418,17 +445,13 @@ type PersistentVolumeClaimSpec struct {
// +optional
VolumeMode *PersistentVolumeMode
// This field can be used to specify either:
// * An existing VolumeSnapshot object (snapshot.storage.k8s.io/VolumeSnapshot - Beta)
// * An existing VolumeSnapshot object (snapshot.storage.k8s.io/VolumeSnapshot)
// * An existing PVC (PersistentVolumeClaim)
// * An existing custom resource/object that implements data population (Alpha)
// In order to use VolumeSnapshot object types, the appropriate feature gate
// must be enabled (VolumeSnapshotDataSource or AnyVolumeDataSource)
// * An existing custom resource that implements data population (Alpha)
// In order to use custom resource types that implement data population,
// the AnyVolumeDataSource feature gate must be enabled.
// If the provisioner or an external controller can support the specified data source,
// it will create a new volume based on the contents of the specified data source.
// If the specified data source is not supported, the volume will
// not be created and the failure will be reported as an event.
// In the future, we plan to support more data source types and the behavior
// of the provisioner may change.
// +optional
DataSource *TypedLocalObjectReference
}
@ -1670,6 +1693,53 @@ type CSIVolumeSource struct {
NodePublishSecretRef *LocalObjectReference
}
// EphemeralVolumeSource represents an ephemeral volume that is handled by a normal storage driver.
type EphemeralVolumeSource struct {
// VolumeClaimTemplate will be used to create a stand-alone PVC to provision the volume.
// The pod in which this EphemeralVolumeSource is embedded will be the
// owner of the PVC, i.e. the PVC will be deleted together with the
// pod. The name of the PVC will be `<pod name>-<volume name>` where
// `<volume name>` is the name from the `PodSpec.Volumes` array
// entry. Pod validation will reject the pod if the concatenated name
// is not valid for a PVC (for example, too long).
//
// An existing PVC with that name that is not owned by the pod
// will *not* be used for the pod to avoid using an unrelated
// volume by mistake. Starting the pod is then blocked until
// the unrelated PVC is removed. If such a pre-created PVC is
// meant to be used by the pod, the PVC has to updated with an
// owner reference to the pod once the pod exists. Normally
// this should not be necessary, but it may be useful when
// manually reconstructing a broken cluster.
//
// This field is read-only and no changes will be made by Kubernetes
// to the PVC after it has been created.
//
// Required, must not be nil.
VolumeClaimTemplate *PersistentVolumeClaimTemplate
// ReadOnly specifies a read-only configuration for the volume.
// Defaults to false (read/write).
// +optional
ReadOnly bool
}
// PersistentVolumeClaimTemplate is used to produce
// PersistentVolumeClaim objects as part of an EphemeralVolumeSource.
type PersistentVolumeClaimTemplate struct {
// ObjectMeta may contain labels and annotations that will be copied into the PVC
// when creating it. No other fields are allowed and will be rejected during
// validation.
// +optional
metav1.ObjectMeta
// Spec for the PersistentVolumeClaim. The entire content is
// copied unchanged into the PVC that gets created from this
// template. The same fields as in a PersistentVolumeClaim
// are also valid here.
Spec PersistentVolumeClaimSpec
}
// ContainerPort represents a network port in a single container
type ContainerPort struct {
// Optional: If specified, this must be an IANA_SVC_NAME Each named port
@ -1774,7 +1844,7 @@ type EnvVar struct {
// EnvVarSource represents a source for the value of an EnvVar.
// Only one of its fields may be set.
type EnvVarSource struct {
// Selects a field of the pod: supports metadata.name, metadata.namespace, metadata.labels, metadata.annotations,
// Selects a field of the pod: supports metadata.name, metadata.namespace, `metadata.labels['<KEY>']`, `metadata.annotations['<KEY>']`,
// metadata.uid, spec.nodeName, spec.serviceAccountName, status.hostIP, status.podIP, status.podIPs.
// +optional
FieldRef *ObjectFieldSelector
@ -2693,6 +2763,12 @@ type PodSpec struct {
// If not specified, the pod will not have a domainname at all.
// +optional
Subdomain string
// If true the pod's hostname will be configured as the pod's FQDN, rather than the leaf name (the default).
// In Linux containers, this means setting the FQDN in the hostname field of the kernel (the nodename field of struct utsname).
// In Windows containers, this means setting the registry value of hostname for the registry key HKEY_LOCAL_MACHINE\\SYSTEM\\CurrentControlSet\\Services\\Tcpip\\Parameters to FQDN.
// If a pod does not have FQDN, this has no effect.
// +optional
SetHostnameAsFQDN *bool
// If specified, the pod's scheduling constraints
// +optional
Affinity *Affinity
@ -2725,7 +2801,7 @@ type PodSpec struct {
// PreemptionPolicy is the Policy for preempting pods with lower priority.
// One of Never, PreemptLowerPriority.
// Defaults to PreemptLowerPriority if unset.
// This field is alpha-level and is only honored by servers that enable the NonPreemptingPriority feature.
// This field is beta-level, gated by the NonPreemptingPriority feature-gate.
// +optional
PreemptionPolicy *PreemptionPolicy
// Specifies the DNS parameters of a pod.
@ -2743,8 +2819,7 @@ type PodSpec struct {
// to run this pod. If no RuntimeClass resource matches the named class, the pod will not be run.
// If unset or empty, the "legacy" RuntimeClass will be used, which is an implicit class with an
// empty definition that uses the default runtime handler.
// More info: https://git.k8s.io/enhancements/keps/sig-node/runtime-class.md
// This is a beta feature as of Kubernetes v1.14.
// More info: https://git.k8s.io/enhancements/keps/sig-node/585-runtime-class/README.md
// +optional
RuntimeClassName *string
// Overhead represents the resource overhead associated with running a pod for a given RuntimeClass.
@ -2764,7 +2839,6 @@ type PodSpec struct {
EnableServiceLinks *bool
// TopologySpreadConstraints describes how a group of pods ought to spread across topology
// domains. Scheduler will schedule pods in a way which abides by the constraints.
// This field is only honored by clusters that enable the EvenPodsSpread feature.
// All topologySpreadConstraints are ANDed.
// +optional
TopologySpreadConstraints []TopologySpreadConstraint
@ -2885,15 +2959,43 @@ type PodSecurityContext struct {
// volume types which support fsGroup based ownership(and permissions).
// It will have no effect on ephemeral volume types such as: secret, configmaps
// and emptydir.
// Valid values are "OnRootMismatch" and "Always". If not specified defaults to "Always".
// Valid values are "OnRootMismatch" and "Always". If not specified, "Always" is used.
// +optional
FSGroupChangePolicy *PodFSGroupChangePolicy
// Sysctls hold a list of namespaced sysctls used for the pod. Pods with unsupported
// sysctls (by the container runtime) might fail to launch.
// +optional
Sysctls []Sysctl
// The seccomp options to use by the containers in this pod.
// +optional
SeccompProfile *SeccompProfile
}
// SeccompProfile defines a pod/container's seccomp profile settings.
// Only one profile source may be set.
// +union
type SeccompProfile struct {
// +unionDiscriminator
Type SeccompProfileType
// Load a profile defined in static file on the node.
// The profile must be preconfigured on the node to work.
// LocalhostProfile cannot be an absolute nor a descending path.
// +optional
LocalhostProfile *string
}
// SeccompProfileType defines the supported seccomp profile types.
type SeccompProfileType string
const (
// SeccompProfileTypeUnconfined is when no seccomp profile is applied (A.K.A. unconfined).
SeccompProfileTypeUnconfined SeccompProfileType = "Unconfined"
// SeccompProfileTypeRuntimeDefault represents the default container runtime seccomp profile.
SeccompProfileTypeRuntimeDefault SeccompProfileType = "RuntimeDefault"
// SeccompProfileTypeLocalhost represents custom made profiles stored on the node's disk.
SeccompProfileTypeLocalhost SeccompProfileType = "Localhost"
)
// PodQOSClass defines the supported qos classes of Pods.
type PodQOSClass string
@ -3373,12 +3475,23 @@ const (
ServiceExternalTrafficPolicyTypeCluster ServiceExternalTrafficPolicyType = "Cluster"
)
// These are the valid conditions of a service.
const (
// LoadBalancerPortsError represents the condition of the requested ports
// on the cloud load balancer instance.
LoadBalancerPortsError = "LoadBalancerPortsError"
)
// ServiceStatus represents the current status of a service
type ServiceStatus struct {
// LoadBalancer contains the current status of the load-balancer,
// if one is present.
// +optional
LoadBalancer LoadBalancerStatus
// Current service condition
// +optional
Conditions []metav1.Condition
}
// LoadBalancerStatus represents the status of a load-balancer
@ -3401,10 +3514,20 @@ type LoadBalancerIngress struct {
// (typically AWS load-balancers)
// +optional
Hostname string
// Ports is a list of records of service ports
// If used, every port defined in the service should have an entry in it
// +optional
Ports []PortStatus
}
const (
// MaxServiceTopologyKeys is the largest number of topology keys allowed on a service
MaxServiceTopologyKeys = 16
)
// IPFamily represents the IP Family (IPv4 or IPv6). This type is used
// to express the family of an IP expressed by a type (i.e. service.Spec.IPFamily)
// to express the family of an IP expressed by a type (e.g. service.spec.ipFamilies).
type IPFamily string
const (
@ -3412,8 +3535,29 @@ const (
IPv4Protocol IPFamily = "IPv4"
// IPv6Protocol indicates that this IP is IPv6 protocol
IPv6Protocol IPFamily = "IPv6"
// MaxServiceTopologyKeys is the largest number of topology keys allowed on a service
MaxServiceTopologyKeys = 16
)
// IPFamilyPolicyType represents the dual-stack-ness requested or required by a Service
type IPFamilyPolicyType string
const (
// IPFamilyPolicySingleStack indicates that this service is required to have a single IPFamily.
// The IPFamily assigned is based on the default IPFamily used by the cluster
// or as identified by service.spec.ipFamilies field
IPFamilyPolicySingleStack IPFamilyPolicyType = "SingleStack"
// IPFamilyPolicyPreferDualStack indicates that this service prefers dual-stack when
// the cluster is configured for dual-stack. If the cluster is not configured
// for dual-stack the service will be assigned a single IPFamily. If the IPFamily is not
// set in service.spec.ipFamilies then the service will be assigned the default IPFamily
// configured on the cluster
IPFamilyPolicyPreferDualStack IPFamilyPolicyType = "PreferDualStack"
// IPFamilyPolicyRequireDualStack indicates that this service requires dual-stack. Using
// IPFamilyPolicyRequireDualStack on a single stack cluster will result in validation errors. The
// IPFamilies (and their order) assigned to this service is based on service.spec.ipFamilies. If
// service.spec.ipFamilies was not provided then it will be assigned according to how they are
// configured on the cluster. If service.spec.ipFamilies has only one entry then the alternative
// IPFamily will be added by apiserver
IPFamilyPolicyRequireDualStack IPFamilyPolicyType = "RequireDualStack"
)
// ServiceSpec describes the attributes that a user creates on a service
@ -3458,6 +3602,36 @@ type ServiceSpec struct {
// +optional
ClusterIP string
// ClusterIPs identifies all the ClusterIPs assigned to this
// service. ClusterIPs are assigned or reserved based on the values of
// service.spec.ipFamilies. A maximum of two entries (dual-stack IPs) are
// allowed in ClusterIPs. The IPFamily of each ClusterIP must match
// values provided in service.spec.ipFamilies. Clients using ClusterIPs must
// keep it in sync with ClusterIP (if provided) by having ClusterIP matching
// first element of ClusterIPs.
// +optional
ClusterIPs []string
// IPFamilies identifies all the IPFamilies assigned for this Service. If a value
// was not provided for IPFamilies it will be defaulted based on the cluster
// configuration and the value of service.spec.ipFamilyPolicy. A maximum of two
// values (dual-stack IPFamilies) are allowed in IPFamilies. IPFamilies field is
// conditionally mutable: it allows for adding or removing a secondary IPFamily,
// but it does not allow changing the primary IPFamily of the service.
// +optional
IPFamilies []IPFamily
// IPFamilyPolicy represents the dual-stack-ness requested or required by this
// Service. If there is no value provided, then this Service will be considered
// SingleStack (single IPFamily). Services can be SingleStack (single IPFamily),
// PreferDualStack (two dual-stack IPFamilies on dual-stack clusters or single
// IPFamily on single-stack clusters), or RequireDualStack (two dual-stack IPFamilies
// on dual-stack configured clusters, otherwise fail). The IPFamilies and ClusterIPs assigned
// to this service can be controlled by service.spec.ipFamilies and service.spec.clusterIPs
// respectively.
// +optional
IPFamilyPolicy *IPFamilyPolicyType
// ExternalName is the external reference that kubedns or equivalent will
// return as a CNAME record for this service. No proxying will be involved.
// Must be a valid RFC-1123 hostname (https://tools.ietf.org/html/rfc1123)
@ -3508,25 +3682,17 @@ type ServiceSpec struct {
// +optional
HealthCheckNodePort int32
// publishNotReadyAddresses, when set to true, indicates that DNS implementations
// must publish the notReadyAddresses of subsets for the Endpoints associated with
// the Service. The default value is false.
// The primary use case for setting this field is to use a StatefulSet's Headless Service
// to propagate SRV records for its Pods without respect to their readiness for purpose
// of peer discovery.
// publishNotReadyAddresses indicates that any agent which deals with endpoints for this
// Service should disregard any indications of ready/not-ready.
// The primary use case for setting this field is for a StatefulSet's Headless Service to
// propagate SRV DNS records for its Pods for the purpose of peer discovery.
// The Kubernetes controllers that generate Endpoints and EndpointSlice resources for
// Services interpret this to mean that all endpoints are considered "ready" even if the
// Pods themselves are not. Agents which consume only Kubernetes generated endpoints
// through the Endpoints or EndpointSlice resources can safely assume this behavior.
// +optional
PublishNotReadyAddresses bool
// ipFamily specifies whether this Service has a preference for a particular IP family (e.g. IPv4 vs.
// IPv6). If a specific IP family is requested, the clusterIP field will be allocated from that family, if it is
// available in the cluster. If no IP family is requested, the cluster's primary IP family will be used.
// Other IP fields (loadBalancerIP, loadBalancerSourceRanges, externalIPs) and controllers which
// allocate external load-balancers should use the same IP family. Endpoints for this Service will be of
// this family. This field is immutable after creation. Assigning a ServiceIPFamily not available in the
// cluster (e.g. IPv6 in IPv4 only cluster) is an error condition and will fail during clusterIP assignment.
// +optional
IPFamily *IPFamily
// topologyKeys is a preference-order list of topology keys which
// implementations of services should use to preferentially sort endpoints
// when accessing this Service, it can not be used at the same time as
@ -3539,8 +3705,18 @@ type ServiceSpec struct {
// The special value "*" may be used to mean "any topology". This catch-all
// value, if used, only makes sense as the last value in the list.
// If this is not specified or empty, no topology constraints will be applied.
// This field is alpha-level and is only honored by servers that enable the ServiceTopology feature.
// +optional
TopologyKeys []string
// allocateLoadBalancerNodePorts defines if NodePorts will be automatically
// allocated for services with type LoadBalancer. Default is "true". It may be
// set to "false" if the cluster load-balancer does not rely on NodePorts.
// allocateLoadBalancerNodePorts may only be set for services with type LoadBalancer
// and will be cleared if the type is changed to any other type.
// This field is alpha-level and is only honored by servers that enable the ServiceLBNodePortControl feature.
// +optional
AllocateLoadBalancerNodePorts *bool
}
// ServicePort represents the port on which the service is exposed
@ -3560,7 +3736,6 @@ type ServicePort struct {
// RFC-6335 and http://www.iana.org/assignments/service-names).
// Non-standard protocols should use prefixed names such as
// mycompany.com/my-custom-protocol.
// Field can be enabled with ServiceAppProtocol feature gate.
// +optional
AppProtocol *string
@ -3711,7 +3886,6 @@ type EndpointPort struct {
// RFC-6335 and http://www.iana.org/assignments/service-names).
// Non-standard protocols should use prefixed names such as
// mycompany.com/my-custom-protocol.
// Field can be enabled with ServiceAppProtocol feature gate.
// +optional
AppProtocol *string
}
@ -3816,7 +3990,7 @@ type NodeSystemInfo struct {
MachineID string
// SystemUUID reported by the node. For unique machine identification
// MachineID is preferred. This field is specific to Red Hat hosts
// https://access.redhat.com/documentation/en-US/Red_Hat_Subscription_Management/1/html/RHSM/getting-system-uuid.html
// https://access.redhat.com/documentation/en-us/red_hat_subscription_management/1/html/rhsm/uuid
SystemUUID string
// Boot ID reported by the node.
BootID string
@ -4037,7 +4211,7 @@ type NodeAddress struct {
}
// NodeResources is an object for conveying resource information about a node.
// see http://releases.k8s.io/HEAD/docs/design/resources.md for more details.
// see https://kubernetes.io/docs/concepts/architecture/nodes/#capacity for more details.
type NodeResources struct {
// Capacity represents the available resources of a node
// +optional
@ -4445,14 +4619,19 @@ const (
// +k8s:deepcopy-gen:interfaces=k8s.io/apimachinery/pkg/runtime.Object
// Event is a report of an event somewhere in the cluster.
// Event is a report of an event somewhere in the cluster. Events
// have a limited retention time and triggers and messages may evolve
// with time. Event consumers should not rely on the timing of an event
// with a given Reason reflecting a consistent underlying trigger, or the
// continued existence of events with that Reason. Events should be
// treated as informative, best-effort, supplemental data.
// TODO: Decide whether to store these separately or with the object they apply to.
type Event struct {
metav1.TypeMeta
// +optional
metav1.ObjectMeta
// Required. The object that this event is about. Mapped to events.Event.regarding
// The object that this event is about. Mapped to events.Event.regarding
// +optional
InvolvedObject ObjectReference
@ -4519,21 +4698,8 @@ type EventSeries struct {
Count int32
// Time of the last occurrence observed
LastObservedTime metav1.MicroTime
// State of this Series: Ongoing or Finished
// Deprecated. Planned removal for 1.18
State EventSeriesState
}
// EventSeriesState defines the state of event series
type EventSeriesState string
// These are valid values of event series state
const (
EventSeriesStateOngoing EventSeriesState = "Ongoing"
EventSeriesStateFinished EventSeriesState = "Finished"
EventSeriesStateUnknown EventSeriesState = "Unknown"
)
// +k8s:deepcopy-gen:interfaces=k8s.io/apimachinery/pkg/runtime.Object
// EventList is a list of events.
@ -4778,7 +4944,7 @@ type Secret struct {
// Immutable field, if set, ensures that data stored in the Secret cannot
// be updated (only object metadata can be modified).
// This is an alpha field enabled by ImmutableEphemeralVolumes feature gate.
// This is a beta field enabled by ImmutableEphemeralVolumes feature gate.
// +optional
Immutable *bool
@ -4787,7 +4953,7 @@ type Secret struct {
// base64 encoded string, representing the arbitrary (possibly non-string)
// data value here.
// +optional
Data map[string][]byte
Data map[string][]byte `datapolicy:"password,security-key,token"`
// Used to facilitate programmatic handling of secret data.
// +optional
@ -4906,7 +5072,7 @@ type ConfigMap struct {
// Immutable field, if set, ensures that data stored in the ConfigMap cannot
// be updated (only object metadata can be modified).
// This is an alpha field enabled by ImmutableEphemeralVolumes feature gate.
// This is a beta field enabled by ImmutableEphemeralVolumes feature gate.
// +optional
Immutable *bool
@ -5003,6 +5169,7 @@ type ComponentCondition struct {
// +k8s:deepcopy-gen:interfaces=k8s.io/apimachinery/pkg/runtime.Object
// ComponentStatus (and ComponentStatusList) holds the cluster validation info.
// Deprecated: This API is deprecated in v1.19+
type ComponentStatus struct {
metav1.TypeMeta
// +optional
@ -5015,6 +5182,7 @@ type ComponentStatus struct {
// +k8s:deepcopy-gen:interfaces=k8s.io/apimachinery/pkg/runtime.Object
// ComponentStatusList represents the list of component statuses
// Deprecated: This API is deprecated in v1.19+
type ComponentStatusList struct {
metav1.TypeMeta
// +optional
@ -5081,6 +5249,11 @@ type SecurityContext struct {
// readonly paths and masked paths.
// +optional
ProcMount *ProcMountType
// The seccomp options to use by this container. If seccomp options are
// provided at both the pod & container level, the container options
// override the pod options.
// +optional
SeccompProfile *SeccompProfile
}
// ProcMountType defines the type of proc mount
@ -5187,8 +5360,8 @@ const (
// TopologySpreadConstraint specifies how to spread matching pods among the given topology.
type TopologySpreadConstraint struct {
// MaxSkew describes the degree to which pods may be unevenly distributed.
// It's the maximum permitted difference between the number of matching pods in
// any two topology domains of a given topology type.
// When `whenUnsatisfiable=DoNotSchedule`, it is the maximum permitted difference
// between the number of matching pods in the target topology and the global minimum.
// For example, in a 3-zone cluster, MaxSkew is set to 1, and pods with the same
// labelSelector spread as 1/1/0:
// +-------+-------+-------+
@ -5200,6 +5373,8 @@ type TopologySpreadConstraint struct {
// scheduling it onto zone1(zone2) would make the ActualSkew(2-0) on zone1(zone2)
// violate MaxSkew(1).
// - if MaxSkew is 2, incoming pod can be scheduled onto any zone.
// When `whenUnsatisfiable=ScheduleAnyway`, it is used to give higher precedence
// to topologies that satisfy it.
// It's a required field. Default value is 1 and 0 is not allowed.
MaxSkew int32
// TopologyKey is the key of node labels. Nodes that have a label with this key
@ -5210,10 +5385,13 @@ type TopologySpreadConstraint struct {
TopologyKey string
// WhenUnsatisfiable indicates how to deal with a pod if it doesn't satisfy
// the spread constraint.
// - DoNotSchedule (default) tells the scheduler not to schedule it
// - ScheduleAnyway tells the scheduler to still schedule it
// It's considered as "Unsatisfiable" if and only if placing incoming pod on any
// topology violates "MaxSkew".
// - DoNotSchedule (default) tells the scheduler not to schedule it.
// - ScheduleAnyway tells the scheduler to schedule the pod in any location,
// but giving higher precedence to topologies that would help reduce the
// skew.
// A constraint is considered "Unsatisfiable" for an incoming pod
// if and only if every possible node assigment for that pod would violate
// "MaxSkew" on some topology.
// For example, in a 3-zone cluster, MaxSkew is set to 1, and pods with the same
// labelSelector spread as 3/1/1:
// +-------+-------+-------+
@ -5233,3 +5411,32 @@ type TopologySpreadConstraint struct {
// +optional
LabelSelector *metav1.LabelSelector
}
// These are the built-in errors for PortStatus.
const (
// MixedProtocolNotSupported error in PortStatus means that the cloud provider
// can't ensure the port on the load balancer because mixed values of protocols
// on the same LoadBalancer type of Service are not supported by the cloud provider.
MixedProtocolNotSupported = "MixedProtocolNotSupported"
)
// PortStatus represents the error condition of a service port
type PortStatus struct {
// Port is the port number of the service port of which status is recorded here
Port int32
// Protocol is the protocol of the service port of which status is recorded here
Protocol Protocol
// Error is to record the problem with the service port
// The format of the error shall comply with the following rules:
// - built-in error values shall be specified in this file and those shall use
// CamelCase names
// - cloud provider specific error values must have names that comply with the
// format foo.example.com/CamelCase.
// ---
// The regex it matches is (dns1123SubdomainFmt/)?(qualifiedNameFmt)
// +optional
// +kubebuilder:validation:Required
// +kubebuilder:validation:Pattern=`^([a-z0-9]([-a-z0-9]*[a-z0-9])?(\.[a-z0-9]([-a-z0-9]*[a-z0-9])?)*/)?(([A-Za-z0-9][-A-Za-z0-9_.]*)?[A-Za-z0-9])$`
// +kubebuilder:validation:MaxLength=316
Error *string
}

View File

@ -27,7 +27,6 @@ go_library(
"//staging/src/k8s.io/apimachinery/pkg/util/intstr:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/validation/field:go_default_library",
"//staging/src/k8s.io/apiserver/pkg/util/feature:go_default_library",
"//vendor/k8s.io/utils/net:go_default_library",
"//vendor/k8s.io/utils/pointer:go_default_library",
],
)

View File

@ -21,13 +21,11 @@ reviewers:
- sttts
- dchen1107
- saad-ali
- zmerlynn
- luxas
- janetkuo
- justinsb
- ncdc
- tallclair
- eparis
- piosz
- jsafrane
- dims

View File

@ -435,6 +435,7 @@ func AddFieldLabelConversionsForEvent(scheme *runtime.Scheme) error {
"involvedObject.resourceVersion",
"involvedObject.fieldPath",
"reason",
"reportingComponent",
"source",
"type",
"metadata.namespace",

View File

@ -19,7 +19,7 @@ package v1
import (
"time"
"k8s.io/api/core/v1"
v1 "k8s.io/api/core/v1"
"k8s.io/apimachinery/pkg/runtime"
"k8s.io/apimachinery/pkg/util/intstr"
"k8s.io/kubernetes/pkg/util/parsers"
@ -27,7 +27,6 @@ import (
utilfeature "k8s.io/apiserver/pkg/util/feature"
"k8s.io/kubernetes/pkg/features"
utilnet "k8s.io/utils/net"
)
func addDefaultingFuncs(scheme *runtime.Scheme) error {
@ -71,11 +70,6 @@ func SetDefaults_Volume(obj *v1.Volume) {
}
}
}
func SetDefaults_ContainerPort(obj *v1.ContainerPort) {
if obj.Protocol == "" {
obj.Protocol = v1.ProtocolTCP
}
}
func SetDefaults_Container(obj *v1.Container) {
if obj.ImagePullPolicy == "" {
// Ignore error and assume it has been validated elsewhere
@ -96,6 +90,10 @@ func SetDefaults_Container(obj *v1.Container) {
}
}
func SetDefaults_EphemeralContainer(obj *v1.EphemeralContainer) {
SetDefaults_Container((*v1.Container)(&obj.EphemeralContainerCommon))
}
func SetDefaults_Service(obj *v1.Service) {
if obj.Spec.SessionAffinity == "" {
obj.Spec.SessionAffinity = v1.ServiceAffinityNone
@ -133,32 +131,44 @@ func SetDefaults_Service(obj *v1.Service) {
obj.Spec.ExternalTrafficPolicy = v1.ServiceExternalTrafficPolicyTypeCluster
}
// if dualstack feature gate is on then we need to default
// Spec.IPFamily correctly. This is to cover the case
// when an existing cluster have been converted to dualstack
// i.e. it already contain services with Spec.IPFamily==nil
if utilfeature.DefaultFeatureGate.Enabled(features.IPv6DualStack) &&
obj.Spec.Type != v1.ServiceTypeExternalName &&
obj.Spec.ClusterIP != "" && /*has an ip already set*/
obj.Spec.ClusterIP != "None" && /* not converting from ExternalName to other */
obj.Spec.IPFamily == nil /* family was not previously set */ {
if utilfeature.DefaultFeatureGate.Enabled(features.IPv6DualStack) {
// Default obj.Spec.IPFamilyPolicy if we *know* we can, otherwise it will
// be handled later in allocation.
if obj.Spec.Type != v1.ServiceTypeExternalName {
if obj.Spec.IPFamilyPolicy == nil {
if len(obj.Spec.ClusterIPs) == 2 || len(obj.Spec.IPFamilies) == 2 {
requireDualStack := v1.IPFamilyPolicyRequireDualStack
obj.Spec.IPFamilyPolicy = &requireDualStack
}
}
// there is a change that the ClusterIP (set by user) is unparsable.
// in this case, the family will be set mistakenly to ipv4 (because
// the util function does not parse errors *sigh*). The error
// will be caught in validation which asserts the validity of the
// IP and the service object will not be persisted with the wrong IP
// family
// If the user demanded dual-stack, but only specified one family, we add
// the other.
if obj.Spec.IPFamilyPolicy != nil && *(obj.Spec.IPFamilyPolicy) == v1.IPFamilyPolicyRequireDualStack && len(obj.Spec.IPFamilies) == 1 {
if obj.Spec.IPFamilies[0] == v1.IPv4Protocol {
obj.Spec.IPFamilies = append(obj.Spec.IPFamilies, v1.IPv6Protocol)
} else {
obj.Spec.IPFamilies = append(obj.Spec.IPFamilies, v1.IPv4Protocol)
}
ipv6 := v1.IPv6Protocol
ipv4 := v1.IPv4Protocol
if utilnet.IsIPv6String(obj.Spec.ClusterIP) {
obj.Spec.IPFamily = &ipv6
} else {
obj.Spec.IPFamily = &ipv4
// Any other dual-stack defaulting depends on cluster configuration.
// Further IPFamilies, IPFamilyPolicy defaulting is in ClusterIP alloc/reserve logic
// NOTE: strategy handles cases where ClusterIPs is used (but not ClusterIP).
}
}
// any other defaulting depends on cluster configuration.
// further IPFamilies, IPFamilyPolicy defaulting is in ClusterIP alloc/reserve logic
// note: conversion logic handles cases where ClusterIPs is used (but not ClusterIP).
}
if utilfeature.DefaultFeatureGate.Enabled(features.ServiceLBNodePortControl) {
if obj.Spec.Type == v1.ServiceTypeLoadBalancer {
if obj.Spec.AllocateLoadBalancerNodePorts == nil {
obj.Spec.AllocateLoadBalancerNodePorts = utilpointer.BoolPtr(true)
}
}
}
}
func SetDefaults_Pod(obj *v1.Pod) {
// If limits are specified, but requests are not, default requests to limits
@ -285,9 +295,11 @@ func SetDefaults_PersistentVolumeClaim(obj *v1.PersistentVolumeClaim) {
if obj.Status.Phase == "" {
obj.Status.Phase = v1.ClaimPending
}
if obj.Spec.VolumeMode == nil {
obj.Spec.VolumeMode = new(v1.PersistentVolumeMode)
*obj.Spec.VolumeMode = v1.PersistentVolumeFilesystem
}
func SetDefaults_PersistentVolumeClaimSpec(obj *v1.PersistentVolumeClaimSpec) {
if obj.VolumeMode == nil {
obj.VolumeMode = new(v1.PersistentVolumeMode)
*obj.VolumeMode = v1.PersistentVolumeFilesystem
}
}
func SetDefaults_ISCSIVolumeSource(obj *v1.ISCSIVolumeSource) {

View File

@ -12,10 +12,8 @@ go_test(
embed = [":go_default_library"],
deps = [
"//staging/src/k8s.io/api/core/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/api/equality:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/api/resource:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/apis/meta/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/fields:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/labels:go_default_library",
],
)
@ -28,7 +26,6 @@ go_library(
"//pkg/apis/core/helper:go_default_library",
"//staging/src/k8s.io/api/core/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/api/resource:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/fields:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/labels:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/selection:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/validation:go_default_library",

View File

@ -23,7 +23,6 @@ import (
v1 "k8s.io/api/core/v1"
"k8s.io/apimachinery/pkg/api/resource"
"k8s.io/apimachinery/pkg/fields"
"k8s.io/apimachinery/pkg/labels"
"k8s.io/apimachinery/pkg/selection"
"k8s.io/apimachinery/pkg/util/validation"
@ -131,22 +130,19 @@ func IsOvercommitAllowed(name v1.ResourceName) bool {
!IsHugePageResourceName(name)
}
// IsAttachableVolumeResourceName returns true when the resource name is prefixed in attachable volume
func IsAttachableVolumeResourceName(name v1.ResourceName) bool {
return strings.HasPrefix(string(name), v1.ResourceAttachableVolumesPrefix)
}
// Extended and Hugepages resources
func IsScalarResourceName(name v1.ResourceName) bool {
return IsExtendedResourceName(name) || IsHugePageResourceName(name) ||
IsPrefixedNativeResource(name) || IsAttachableVolumeResourceName(name)
}
// this function aims to check if the service's ClusterIP is set or not
// IsServiceIPSet aims to check if the service's ClusterIP is set or not
// the objective is not to perform validation here
func IsServiceIPSet(service *v1.Service) bool {
return service.Spec.ClusterIP != v1.ClusterIPNone && service.Spec.ClusterIP != ""
}
// LoadBalancerStatusEqual evaluates the given load balancers' ingress IP addresses
// and hostnames and returns true if equal or false if otherwise
// TODO: make method on LoadBalancerStatus?
func LoadBalancerStatusEqual(l, r *v1.LoadBalancerStatus) bool {
return ingressSliceEqual(l.Ingress, r.Ingress)
@ -191,7 +187,7 @@ func GetAccessModesAsString(modes []v1.PersistentVolumeAccessMode) string {
return strings.Join(modesStr, ",")
}
// GetAccessModesAsString returns an array of AccessModes from a string created by GetAccessModesAsString
// GetAccessModesFromString returns an array of AccessModes from a string created by GetAccessModesAsString
func GetAccessModesFromString(modes string) []v1.PersistentVolumeAccessMode {
strmodes := strings.Split(modes, ",")
accessModes := []v1.PersistentVolumeAccessMode{}
@ -229,72 +225,6 @@ func containsAccessMode(modes []v1.PersistentVolumeAccessMode, mode v1.Persisten
return false
}
// NodeSelectorRequirementsAsSelector converts the []NodeSelectorRequirement api type into a struct that implements
// labels.Selector.
func NodeSelectorRequirementsAsSelector(nsm []v1.NodeSelectorRequirement) (labels.Selector, error) {
if len(nsm) == 0 {
return labels.Nothing(), nil
}
selector := labels.NewSelector()
for _, expr := range nsm {
var op selection.Operator
switch expr.Operator {
case v1.NodeSelectorOpIn:
op = selection.In
case v1.NodeSelectorOpNotIn:
op = selection.NotIn
case v1.NodeSelectorOpExists:
op = selection.Exists
case v1.NodeSelectorOpDoesNotExist:
op = selection.DoesNotExist
case v1.NodeSelectorOpGt:
op = selection.GreaterThan
case v1.NodeSelectorOpLt:
op = selection.LessThan
default:
return nil, fmt.Errorf("%q is not a valid node selector operator", expr.Operator)
}
r, err := labels.NewRequirement(expr.Key, op, expr.Values)
if err != nil {
return nil, err
}
selector = selector.Add(*r)
}
return selector, nil
}
// NodeSelectorRequirementsAsFieldSelector converts the []NodeSelectorRequirement core type into a struct that implements
// fields.Selector.
func NodeSelectorRequirementsAsFieldSelector(nsm []v1.NodeSelectorRequirement) (fields.Selector, error) {
if len(nsm) == 0 {
return fields.Nothing(), nil
}
selectors := []fields.Selector{}
for _, expr := range nsm {
switch expr.Operator {
case v1.NodeSelectorOpIn:
if len(expr.Values) != 1 {
return nil, fmt.Errorf("unexpected number of value (%d) for node field selector operator %q",
len(expr.Values), expr.Operator)
}
selectors = append(selectors, fields.OneTermEqualSelector(expr.Key, expr.Values[0]))
case v1.NodeSelectorOpNotIn:
if len(expr.Values) != 1 {
return nil, fmt.Errorf("unexpected number of value (%d) for node field selector operator %q",
len(expr.Values), expr.Operator)
}
selectors = append(selectors, fields.OneTermNotEqualSelector(expr.Key, expr.Values[0]))
default:
return nil, fmt.Errorf("%q is not a valid node field selector operator", expr.Operator)
}
}
return fields.AndSelectors(selectors...), nil
}
// NodeSelectorRequirementKeysExistInNodeSelectorTerms checks if a NodeSelectorTerm with key is already specified in terms
func NodeSelectorRequirementKeysExistInNodeSelectorTerms(reqs []v1.NodeSelectorRequirement, terms []v1.NodeSelectorTerm) bool {
for _, req := range reqs {
@ -309,39 +239,6 @@ func NodeSelectorRequirementKeysExistInNodeSelectorTerms(reqs []v1.NodeSelectorR
return false
}
// MatchNodeSelectorTerms checks whether the node labels and fields match node selector terms in ORed;
// nil or empty term matches no objects.
func MatchNodeSelectorTerms(
nodeSelectorTerms []v1.NodeSelectorTerm,
nodeLabels labels.Set,
nodeFields fields.Set,
) bool {
for _, req := range nodeSelectorTerms {
// nil or empty term selects no objects
if len(req.MatchExpressions) == 0 && len(req.MatchFields) == 0 {
continue
}
if len(req.MatchExpressions) != 0 {
labelSelector, err := NodeSelectorRequirementsAsSelector(req.MatchExpressions)
if err != nil || !labelSelector.Matches(nodeLabels) {
continue
}
}
if len(req.MatchFields) != 0 {
fieldSelector, err := NodeSelectorRequirementsAsFieldSelector(req.MatchFields)
if err != nil || !fieldSelector.Matches(nodeFields) {
continue
}
}
return true
}
return false
}
// TopologySelectorRequirementsAsSelector converts the []TopologySelectorLabelRequirement api type into a struct
// that implements labels.Selector.
func TopologySelectorRequirementsAsSelector(tsm []v1.TopologySelectorLabelRequirement) (labels.Selector, error) {
@ -467,7 +364,7 @@ func getFilteredTaints(taints []v1.Taint, inclusionFilter taintsFilterFunc) []v1
return filteredTaints
}
// Returns true and list of Tolerations matching all Taints if all are tolerated, or false otherwise.
// GetMatchingTolerations returns true and list of Tolerations matching all Taints if all are tolerated, or false otherwise.
func GetMatchingTolerations(taints []v1.Taint, tolerations []v1.Toleration) (bool, []v1.Toleration) {
if len(taints) == 0 {
return true, []v1.Toleration{}
@ -492,6 +389,8 @@ func GetMatchingTolerations(taints []v1.Taint, tolerations []v1.Toleration) (boo
return true, result
}
// GetAvoidPodsFromNodeAnnotations scans the list of annotations and
// returns the pods that needs to be avoided for this node from scheduling
func GetAvoidPodsFromNodeAnnotations(annotations map[string]string) (v1.AvoidPods, error) {
var avoidPods v1.AvoidPods
if len(annotations) > 0 && annotations[v1.PreferAvoidPodsAnnotationKey] != "" {

View File

@ -541,6 +541,16 @@ func RegisterConversions(s *runtime.Scheme) error {
}); err != nil {
return err
}
if err := s.AddGeneratedConversionFunc((*v1.EphemeralVolumeSource)(nil), (*core.EphemeralVolumeSource)(nil), func(a, b interface{}, scope conversion.Scope) error {
return Convert_v1_EphemeralVolumeSource_To_core_EphemeralVolumeSource(a.(*v1.EphemeralVolumeSource), b.(*core.EphemeralVolumeSource), scope)
}); err != nil {
return err
}
if err := s.AddGeneratedConversionFunc((*core.EphemeralVolumeSource)(nil), (*v1.EphemeralVolumeSource)(nil), func(a, b interface{}, scope conversion.Scope) error {
return Convert_core_EphemeralVolumeSource_To_v1_EphemeralVolumeSource(a.(*core.EphemeralVolumeSource), b.(*v1.EphemeralVolumeSource), scope)
}); err != nil {
return err
}
if err := s.AddGeneratedConversionFunc((*v1.Event)(nil), (*core.Event)(nil), func(a, b interface{}, scope conversion.Scope) error {
return Convert_v1_Event_To_core_Event(a.(*v1.Event), b.(*core.Event), scope)
}); err != nil {
@ -1131,6 +1141,16 @@ func RegisterConversions(s *runtime.Scheme) error {
}); err != nil {
return err
}
if err := s.AddGeneratedConversionFunc((*v1.PersistentVolumeClaimTemplate)(nil), (*core.PersistentVolumeClaimTemplate)(nil), func(a, b interface{}, scope conversion.Scope) error {
return Convert_v1_PersistentVolumeClaimTemplate_To_core_PersistentVolumeClaimTemplate(a.(*v1.PersistentVolumeClaimTemplate), b.(*core.PersistentVolumeClaimTemplate), scope)
}); err != nil {
return err
}
if err := s.AddGeneratedConversionFunc((*core.PersistentVolumeClaimTemplate)(nil), (*v1.PersistentVolumeClaimTemplate)(nil), func(a, b interface{}, scope conversion.Scope) error {
return Convert_core_PersistentVolumeClaimTemplate_To_v1_PersistentVolumeClaimTemplate(a.(*core.PersistentVolumeClaimTemplate), b.(*v1.PersistentVolumeClaimTemplate), scope)
}); err != nil {
return err
}
if err := s.AddGeneratedConversionFunc((*v1.PersistentVolumeClaimVolumeSource)(nil), (*core.PersistentVolumeClaimVolumeSource)(nil), func(a, b interface{}, scope conversion.Scope) error {
return Convert_v1_PersistentVolumeClaimVolumeSource_To_core_PersistentVolumeClaimVolumeSource(a.(*v1.PersistentVolumeClaimVolumeSource), b.(*core.PersistentVolumeClaimVolumeSource), scope)
}); err != nil {
@ -1371,6 +1391,16 @@ func RegisterConversions(s *runtime.Scheme) error {
}); err != nil {
return err
}
if err := s.AddGeneratedConversionFunc((*v1.PortStatus)(nil), (*core.PortStatus)(nil), func(a, b interface{}, scope conversion.Scope) error {
return Convert_v1_PortStatus_To_core_PortStatus(a.(*v1.PortStatus), b.(*core.PortStatus), scope)
}); err != nil {
return err
}
if err := s.AddGeneratedConversionFunc((*core.PortStatus)(nil), (*v1.PortStatus)(nil), func(a, b interface{}, scope conversion.Scope) error {
return Convert_core_PortStatus_To_v1_PortStatus(a.(*core.PortStatus), b.(*v1.PortStatus), scope)
}); err != nil {
return err
}
if err := s.AddGeneratedConversionFunc((*v1.PortworxVolumeSource)(nil), (*core.PortworxVolumeSource)(nil), func(a, b interface{}, scope conversion.Scope) error {
return Convert_v1_PortworxVolumeSource_To_core_PortworxVolumeSource(a.(*v1.PortworxVolumeSource), b.(*core.PortworxVolumeSource), scope)
}); err != nil {
@ -1621,6 +1651,16 @@ func RegisterConversions(s *runtime.Scheme) error {
}); err != nil {
return err
}
if err := s.AddGeneratedConversionFunc((*v1.SeccompProfile)(nil), (*core.SeccompProfile)(nil), func(a, b interface{}, scope conversion.Scope) error {
return Convert_v1_SeccompProfile_To_core_SeccompProfile(a.(*v1.SeccompProfile), b.(*core.SeccompProfile), scope)
}); err != nil {
return err
}
if err := s.AddGeneratedConversionFunc((*core.SeccompProfile)(nil), (*v1.SeccompProfile)(nil), func(a, b interface{}, scope conversion.Scope) error {
return Convert_core_SeccompProfile_To_v1_SeccompProfile(a.(*core.SeccompProfile), b.(*v1.SeccompProfile), scope)
}); err != nil {
return err
}
if err := s.AddGeneratedConversionFunc((*core.Secret)(nil), (*v1.Secret)(nil), func(a, b interface{}, scope conversion.Scope) error {
return Convert_core_Secret_To_v1_Secret(a.(*core.Secret), b.(*v1.Secret), scope)
}); err != nil {
@ -3504,6 +3544,28 @@ func Convert_core_EphemeralContainers_To_v1_EphemeralContainers(in *core.Ephemer
return autoConvert_core_EphemeralContainers_To_v1_EphemeralContainers(in, out, s)
}
func autoConvert_v1_EphemeralVolumeSource_To_core_EphemeralVolumeSource(in *v1.EphemeralVolumeSource, out *core.EphemeralVolumeSource, s conversion.Scope) error {
out.VolumeClaimTemplate = (*core.PersistentVolumeClaimTemplate)(unsafe.Pointer(in.VolumeClaimTemplate))
out.ReadOnly = in.ReadOnly
return nil
}
// Convert_v1_EphemeralVolumeSource_To_core_EphemeralVolumeSource is an autogenerated conversion function.
func Convert_v1_EphemeralVolumeSource_To_core_EphemeralVolumeSource(in *v1.EphemeralVolumeSource, out *core.EphemeralVolumeSource, s conversion.Scope) error {
return autoConvert_v1_EphemeralVolumeSource_To_core_EphemeralVolumeSource(in, out, s)
}
func autoConvert_core_EphemeralVolumeSource_To_v1_EphemeralVolumeSource(in *core.EphemeralVolumeSource, out *v1.EphemeralVolumeSource, s conversion.Scope) error {
out.VolumeClaimTemplate = (*v1.PersistentVolumeClaimTemplate)(unsafe.Pointer(in.VolumeClaimTemplate))
out.ReadOnly = in.ReadOnly
return nil
}
// Convert_core_EphemeralVolumeSource_To_v1_EphemeralVolumeSource is an autogenerated conversion function.
func Convert_core_EphemeralVolumeSource_To_v1_EphemeralVolumeSource(in *core.EphemeralVolumeSource, out *v1.EphemeralVolumeSource, s conversion.Scope) error {
return autoConvert_core_EphemeralVolumeSource_To_v1_EphemeralVolumeSource(in, out, s)
}
func autoConvert_v1_Event_To_core_Event(in *v1.Event, out *core.Event, s conversion.Scope) error {
out.ObjectMeta = in.ObjectMeta
if err := Convert_v1_ObjectReference_To_core_ObjectReference(&in.InvolvedObject, &out.InvolvedObject, s); err != nil {
@ -3585,7 +3647,6 @@ func Convert_core_EventList_To_v1_EventList(in *core.EventList, out *v1.EventLis
func autoConvert_v1_EventSeries_To_core_EventSeries(in *v1.EventSeries, out *core.EventSeries, s conversion.Scope) error {
out.Count = in.Count
out.LastObservedTime = in.LastObservedTime
out.State = core.EventSeriesState(in.State)
return nil
}
@ -3597,7 +3658,6 @@ func Convert_v1_EventSeries_To_core_EventSeries(in *v1.EventSeries, out *core.Ev
func autoConvert_core_EventSeries_To_v1_EventSeries(in *core.EventSeries, out *v1.EventSeries, s conversion.Scope) error {
out.Count = in.Count
out.LastObservedTime = in.LastObservedTime
out.State = v1.EventSeriesState(in.State)
return nil
}
@ -4241,6 +4301,7 @@ func Convert_core_List_To_v1_List(in *core.List, out *v1.List, s conversion.Scop
func autoConvert_v1_LoadBalancerIngress_To_core_LoadBalancerIngress(in *v1.LoadBalancerIngress, out *core.LoadBalancerIngress, s conversion.Scope) error {
out.IP = in.IP
out.Hostname = in.Hostname
out.Ports = *(*[]core.PortStatus)(unsafe.Pointer(&in.Ports))
return nil
}
@ -4252,6 +4313,7 @@ func Convert_v1_LoadBalancerIngress_To_core_LoadBalancerIngress(in *v1.LoadBalan
func autoConvert_core_LoadBalancerIngress_To_v1_LoadBalancerIngress(in *core.LoadBalancerIngress, out *v1.LoadBalancerIngress, s conversion.Scope) error {
out.IP = in.IP
out.Hostname = in.Hostname
out.Ports = *(*[]v1.PortStatus)(unsafe.Pointer(&in.Ports))
return nil
}
@ -5141,6 +5203,32 @@ func Convert_core_PersistentVolumeClaimStatus_To_v1_PersistentVolumeClaimStatus(
return autoConvert_core_PersistentVolumeClaimStatus_To_v1_PersistentVolumeClaimStatus(in, out, s)
}
func autoConvert_v1_PersistentVolumeClaimTemplate_To_core_PersistentVolumeClaimTemplate(in *v1.PersistentVolumeClaimTemplate, out *core.PersistentVolumeClaimTemplate, s conversion.Scope) error {
out.ObjectMeta = in.ObjectMeta
if err := Convert_v1_PersistentVolumeClaimSpec_To_core_PersistentVolumeClaimSpec(&in.Spec, &out.Spec, s); err != nil {
return err
}
return nil
}
// Convert_v1_PersistentVolumeClaimTemplate_To_core_PersistentVolumeClaimTemplate is an autogenerated conversion function.
func Convert_v1_PersistentVolumeClaimTemplate_To_core_PersistentVolumeClaimTemplate(in *v1.PersistentVolumeClaimTemplate, out *core.PersistentVolumeClaimTemplate, s conversion.Scope) error {
return autoConvert_v1_PersistentVolumeClaimTemplate_To_core_PersistentVolumeClaimTemplate(in, out, s)
}
func autoConvert_core_PersistentVolumeClaimTemplate_To_v1_PersistentVolumeClaimTemplate(in *core.PersistentVolumeClaimTemplate, out *v1.PersistentVolumeClaimTemplate, s conversion.Scope) error {
out.ObjectMeta = in.ObjectMeta
if err := Convert_core_PersistentVolumeClaimSpec_To_v1_PersistentVolumeClaimSpec(&in.Spec, &out.Spec, s); err != nil {
return err
}
return nil
}
// Convert_core_PersistentVolumeClaimTemplate_To_v1_PersistentVolumeClaimTemplate is an autogenerated conversion function.
func Convert_core_PersistentVolumeClaimTemplate_To_v1_PersistentVolumeClaimTemplate(in *core.PersistentVolumeClaimTemplate, out *v1.PersistentVolumeClaimTemplate, s conversion.Scope) error {
return autoConvert_core_PersistentVolumeClaimTemplate_To_v1_PersistentVolumeClaimTemplate(in, out, s)
}
func autoConvert_v1_PersistentVolumeClaimVolumeSource_To_core_PersistentVolumeClaimVolumeSource(in *v1.PersistentVolumeClaimVolumeSource, out *core.PersistentVolumeClaimVolumeSource, s conversion.Scope) error {
out.ClaimName = in.ClaimName
out.ReadOnly = in.ReadOnly
@ -5942,6 +6030,7 @@ func autoConvert_v1_PodSecurityContext_To_core_PodSecurityContext(in *v1.PodSecu
out.FSGroup = (*int64)(unsafe.Pointer(in.FSGroup))
out.Sysctls = *(*[]core.Sysctl)(unsafe.Pointer(&in.Sysctls))
out.FSGroupChangePolicy = (*core.PodFSGroupChangePolicy)(unsafe.Pointer(in.FSGroupChangePolicy))
out.SeccompProfile = (*core.SeccompProfile)(unsafe.Pointer(in.SeccompProfile))
return nil
}
@ -5964,6 +6053,7 @@ func autoConvert_core_PodSecurityContext_To_v1_PodSecurityContext(in *core.PodSe
out.FSGroup = (*int64)(unsafe.Pointer(in.FSGroup))
out.FSGroupChangePolicy = (*v1.PodFSGroupChangePolicy)(unsafe.Pointer(in.FSGroupChangePolicy))
out.Sysctls = *(*[]v1.Sysctl)(unsafe.Pointer(&in.Sysctls))
out.SeccompProfile = (*v1.SeccompProfile)(unsafe.Pointer(in.SeccompProfile))
return nil
}
@ -6045,6 +6135,7 @@ func autoConvert_v1_PodSpec_To_core_PodSpec(in *v1.PodSpec, out *core.PodSpec, s
out.PreemptionPolicy = (*core.PreemptionPolicy)(unsafe.Pointer(in.PreemptionPolicy))
out.Overhead = *(*core.ResourceList)(unsafe.Pointer(&in.Overhead))
out.TopologySpreadConstraints = *(*[]core.TopologySpreadConstraint)(unsafe.Pointer(&in.TopologySpreadConstraints))
out.SetHostnameAsFQDN = (*bool)(unsafe.Pointer(in.SetHostnameAsFQDN))
return nil
}
@ -6083,6 +6174,7 @@ func autoConvert_core_PodSpec_To_v1_PodSpec(in *core.PodSpec, out *v1.PodSpec, s
out.ImagePullSecrets = *(*[]v1.LocalObjectReference)(unsafe.Pointer(&in.ImagePullSecrets))
out.Hostname = in.Hostname
out.Subdomain = in.Subdomain
out.SetHostnameAsFQDN = (*bool)(unsafe.Pointer(in.SetHostnameAsFQDN))
out.Affinity = (*v1.Affinity)(unsafe.Pointer(in.Affinity))
out.SchedulerName = in.SchedulerName
out.Tolerations = *(*[]v1.Toleration)(unsafe.Pointer(&in.Tolerations))
@ -6242,6 +6334,30 @@ func autoConvert_core_PodTemplateSpec_To_v1_PodTemplateSpec(in *core.PodTemplate
return nil
}
func autoConvert_v1_PortStatus_To_core_PortStatus(in *v1.PortStatus, out *core.PortStatus, s conversion.Scope) error {
out.Port = in.Port
out.Protocol = core.Protocol(in.Protocol)
out.Error = (*string)(unsafe.Pointer(in.Error))
return nil
}
// Convert_v1_PortStatus_To_core_PortStatus is an autogenerated conversion function.
func Convert_v1_PortStatus_To_core_PortStatus(in *v1.PortStatus, out *core.PortStatus, s conversion.Scope) error {
return autoConvert_v1_PortStatus_To_core_PortStatus(in, out, s)
}
func autoConvert_core_PortStatus_To_v1_PortStatus(in *core.PortStatus, out *v1.PortStatus, s conversion.Scope) error {
out.Port = in.Port
out.Protocol = v1.Protocol(in.Protocol)
out.Error = (*string)(unsafe.Pointer(in.Error))
return nil
}
// Convert_core_PortStatus_To_v1_PortStatus is an autogenerated conversion function.
func Convert_core_PortStatus_To_v1_PortStatus(in *core.PortStatus, out *v1.PortStatus, s conversion.Scope) error {
return autoConvert_core_PortStatus_To_v1_PortStatus(in, out, s)
}
func autoConvert_v1_PortworxVolumeSource_To_core_PortworxVolumeSource(in *v1.PortworxVolumeSource, out *core.PortworxVolumeSource, s conversion.Scope) error {
out.VolumeID = in.VolumeID
out.FSType = in.FSType
@ -7000,6 +7116,28 @@ func Convert_core_ScopedResourceSelectorRequirement_To_v1_ScopedResourceSelector
return autoConvert_core_ScopedResourceSelectorRequirement_To_v1_ScopedResourceSelectorRequirement(in, out, s)
}
func autoConvert_v1_SeccompProfile_To_core_SeccompProfile(in *v1.SeccompProfile, out *core.SeccompProfile, s conversion.Scope) error {
out.Type = core.SeccompProfileType(in.Type)
out.LocalhostProfile = (*string)(unsafe.Pointer(in.LocalhostProfile))
return nil
}
// Convert_v1_SeccompProfile_To_core_SeccompProfile is an autogenerated conversion function.
func Convert_v1_SeccompProfile_To_core_SeccompProfile(in *v1.SeccompProfile, out *core.SeccompProfile, s conversion.Scope) error {
return autoConvert_v1_SeccompProfile_To_core_SeccompProfile(in, out, s)
}
func autoConvert_core_SeccompProfile_To_v1_SeccompProfile(in *core.SeccompProfile, out *v1.SeccompProfile, s conversion.Scope) error {
out.Type = v1.SeccompProfileType(in.Type)
out.LocalhostProfile = (*string)(unsafe.Pointer(in.LocalhostProfile))
return nil
}
// Convert_core_SeccompProfile_To_v1_SeccompProfile is an autogenerated conversion function.
func Convert_core_SeccompProfile_To_v1_SeccompProfile(in *core.SeccompProfile, out *v1.SeccompProfile, s conversion.Scope) error {
return autoConvert_core_SeccompProfile_To_v1_SeccompProfile(in, out, s)
}
func autoConvert_v1_Secret_To_core_Secret(in *v1.Secret, out *core.Secret, s conversion.Scope) error {
out.ObjectMeta = in.ObjectMeta
out.Immutable = (*bool)(unsafe.Pointer(in.Immutable))
@ -7205,6 +7343,7 @@ func autoConvert_v1_SecurityContext_To_core_SecurityContext(in *v1.SecurityConte
out.ReadOnlyRootFilesystem = (*bool)(unsafe.Pointer(in.ReadOnlyRootFilesystem))
out.AllowPrivilegeEscalation = (*bool)(unsafe.Pointer(in.AllowPrivilegeEscalation))
out.ProcMount = (*core.ProcMountType)(unsafe.Pointer(in.ProcMount))
out.SeccompProfile = (*core.SeccompProfile)(unsafe.Pointer(in.SeccompProfile))
return nil
}
@ -7224,6 +7363,7 @@ func autoConvert_core_SecurityContext_To_v1_SecurityContext(in *core.SecurityCon
out.ReadOnlyRootFilesystem = (*bool)(unsafe.Pointer(in.ReadOnlyRootFilesystem))
out.AllowPrivilegeEscalation = (*bool)(unsafe.Pointer(in.AllowPrivilegeEscalation))
out.ProcMount = (*v1.ProcMountType)(unsafe.Pointer(in.ProcMount))
out.SeccompProfile = (*v1.SeccompProfile)(unsafe.Pointer(in.SeccompProfile))
return nil
}
@ -7478,6 +7618,7 @@ func autoConvert_v1_ServiceSpec_To_core_ServiceSpec(in *v1.ServiceSpec, out *cor
out.Ports = *(*[]core.ServicePort)(unsafe.Pointer(&in.Ports))
out.Selector = *(*map[string]string)(unsafe.Pointer(&in.Selector))
out.ClusterIP = in.ClusterIP
out.ClusterIPs = *(*[]string)(unsafe.Pointer(&in.ClusterIPs))
out.Type = core.ServiceType(in.Type)
out.ExternalIPs = *(*[]string)(unsafe.Pointer(&in.ExternalIPs))
out.SessionAffinity = core.ServiceAffinity(in.SessionAffinity)
@ -7488,8 +7629,10 @@ func autoConvert_v1_ServiceSpec_To_core_ServiceSpec(in *v1.ServiceSpec, out *cor
out.HealthCheckNodePort = in.HealthCheckNodePort
out.PublishNotReadyAddresses = in.PublishNotReadyAddresses
out.SessionAffinityConfig = (*core.SessionAffinityConfig)(unsafe.Pointer(in.SessionAffinityConfig))
out.IPFamily = (*core.IPFamily)(unsafe.Pointer(in.IPFamily))
out.TopologyKeys = *(*[]string)(unsafe.Pointer(&in.TopologyKeys))
out.IPFamilies = *(*[]core.IPFamily)(unsafe.Pointer(&in.IPFamilies))
out.IPFamilyPolicy = (*core.IPFamilyPolicyType)(unsafe.Pointer(in.IPFamilyPolicy))
out.AllocateLoadBalancerNodePorts = (*bool)(unsafe.Pointer(in.AllocateLoadBalancerNodePorts))
return nil
}
@ -7503,6 +7646,9 @@ func autoConvert_core_ServiceSpec_To_v1_ServiceSpec(in *core.ServiceSpec, out *v
out.Ports = *(*[]v1.ServicePort)(unsafe.Pointer(&in.Ports))
out.Selector = *(*map[string]string)(unsafe.Pointer(&in.Selector))
out.ClusterIP = in.ClusterIP
out.ClusterIPs = *(*[]string)(unsafe.Pointer(&in.ClusterIPs))
out.IPFamilies = *(*[]v1.IPFamily)(unsafe.Pointer(&in.IPFamilies))
out.IPFamilyPolicy = (*v1.IPFamilyPolicyType)(unsafe.Pointer(in.IPFamilyPolicy))
out.ExternalName = in.ExternalName
out.ExternalIPs = *(*[]string)(unsafe.Pointer(&in.ExternalIPs))
out.LoadBalancerIP = in.LoadBalancerIP
@ -7512,8 +7658,8 @@ func autoConvert_core_ServiceSpec_To_v1_ServiceSpec(in *core.ServiceSpec, out *v
out.ExternalTrafficPolicy = v1.ServiceExternalTrafficPolicyType(in.ExternalTrafficPolicy)
out.HealthCheckNodePort = in.HealthCheckNodePort
out.PublishNotReadyAddresses = in.PublishNotReadyAddresses
out.IPFamily = (*v1.IPFamily)(unsafe.Pointer(in.IPFamily))
out.TopologyKeys = *(*[]string)(unsafe.Pointer(&in.TopologyKeys))
out.AllocateLoadBalancerNodePorts = (*bool)(unsafe.Pointer(in.AllocateLoadBalancerNodePorts))
return nil
}
@ -7526,6 +7672,7 @@ func autoConvert_v1_ServiceStatus_To_core_ServiceStatus(in *v1.ServiceStatus, ou
if err := Convert_v1_LoadBalancerStatus_To_core_LoadBalancerStatus(&in.LoadBalancer, &out.LoadBalancer, s); err != nil {
return err
}
out.Conditions = *(*[]metav1.Condition)(unsafe.Pointer(&in.Conditions))
return nil
}
@ -7538,6 +7685,7 @@ func autoConvert_core_ServiceStatus_To_v1_ServiceStatus(in *core.ServiceStatus,
if err := Convert_core_LoadBalancerStatus_To_v1_LoadBalancerStatus(&in.LoadBalancer, &out.LoadBalancer, s); err != nil {
return err
}
out.Conditions = *(*[]metav1.Condition)(unsafe.Pointer(&in.Conditions))
return nil
}
@ -7979,6 +8127,7 @@ func autoConvert_v1_VolumeSource_To_core_VolumeSource(in *v1.VolumeSource, out *
out.ScaleIO = (*core.ScaleIOVolumeSource)(unsafe.Pointer(in.ScaleIO))
out.StorageOS = (*core.StorageOSVolumeSource)(unsafe.Pointer(in.StorageOS))
out.CSI = (*core.CSIVolumeSource)(unsafe.Pointer(in.CSI))
out.Ephemeral = (*core.EphemeralVolumeSource)(unsafe.Pointer(in.Ephemeral))
return nil
}
@ -8024,6 +8173,7 @@ func autoConvert_core_VolumeSource_To_v1_VolumeSource(in *core.VolumeSource, out
out.ScaleIO = (*v1.ScaleIOVolumeSource)(unsafe.Pointer(in.ScaleIO))
out.StorageOS = (*v1.StorageOSVolumeSource)(unsafe.Pointer(in.StorageOS))
out.CSI = (*v1.CSIVolumeSource)(unsafe.Pointer(in.CSI))
out.Ephemeral = (*v1.EphemeralVolumeSource)(unsafe.Pointer(in.Ephemeral))
return nil
}

View File

@ -21,6 +21,8 @@ limitations under the License.
package v1
import (
"reflect"
v1 "k8s.io/api/core/v1"
runtime "k8s.io/apimachinery/pkg/runtime"
)
@ -88,9 +90,12 @@ func SetObjectDefaults_EndpointsList(in *v1.EndpointsList) {
func SetObjectDefaults_EphemeralContainers(in *v1.EphemeralContainers) {
for i := range in.EphemeralContainers {
a := &in.EphemeralContainers[i]
SetDefaults_EphemeralContainer(a)
for j := range a.EphemeralContainerCommon.Ports {
b := &a.EphemeralContainerCommon.Ports[j]
SetDefaults_ContainerPort(b)
if reflect.ValueOf(b.Protocol).IsZero() {
b.Protocol = "TCP"
}
}
for j := range a.EphemeralContainerCommon.Env {
b := &a.EphemeralContainerCommon.Env[j]
@ -200,6 +205,7 @@ func SetObjectDefaults_PersistentVolume(in *v1.PersistentVolume) {
func SetObjectDefaults_PersistentVolumeClaim(in *v1.PersistentVolumeClaim) {
SetDefaults_PersistentVolumeClaim(in)
SetDefaults_PersistentVolumeClaimSpec(&in.Spec)
SetDefaults_ResourceList(&in.Spec.Resources.Limits)
SetDefaults_ResourceList(&in.Spec.Resources.Requests)
SetDefaults_ResourceList(&in.Status.Capacity)
@ -272,13 +278,22 @@ func SetObjectDefaults_Pod(in *v1.Pod) {
if a.VolumeSource.ScaleIO != nil {
SetDefaults_ScaleIOVolumeSource(a.VolumeSource.ScaleIO)
}
if a.VolumeSource.Ephemeral != nil {
if a.VolumeSource.Ephemeral.VolumeClaimTemplate != nil {
SetDefaults_PersistentVolumeClaimSpec(&a.VolumeSource.Ephemeral.VolumeClaimTemplate.Spec)
SetDefaults_ResourceList(&a.VolumeSource.Ephemeral.VolumeClaimTemplate.Spec.Resources.Limits)
SetDefaults_ResourceList(&a.VolumeSource.Ephemeral.VolumeClaimTemplate.Spec.Resources.Requests)
}
}
}
for i := range in.Spec.InitContainers {
a := &in.Spec.InitContainers[i]
SetDefaults_Container(a)
for j := range a.Ports {
b := &a.Ports[j]
SetDefaults_ContainerPort(b)
if reflect.ValueOf(b.Protocol).IsZero() {
b.Protocol = "TCP"
}
}
for j := range a.Env {
b := &a.Env[j]
@ -326,7 +341,9 @@ func SetObjectDefaults_Pod(in *v1.Pod) {
SetDefaults_Container(a)
for j := range a.Ports {
b := &a.Ports[j]
SetDefaults_ContainerPort(b)
if reflect.ValueOf(b.Protocol).IsZero() {
b.Protocol = "TCP"
}
}
for j := range a.Env {
b := &a.Env[j]
@ -371,9 +388,12 @@ func SetObjectDefaults_Pod(in *v1.Pod) {
}
for i := range in.Spec.EphemeralContainers {
a := &in.Spec.EphemeralContainers[i]
SetDefaults_EphemeralContainer(a)
for j := range a.EphemeralContainerCommon.Ports {
b := &a.EphemeralContainerCommon.Ports[j]
SetDefaults_ContainerPort(b)
if reflect.ValueOf(b.Protocol).IsZero() {
b.Protocol = "TCP"
}
}
for j := range a.EphemeralContainerCommon.Env {
b := &a.EphemeralContainerCommon.Env[j]
@ -478,13 +498,22 @@ func SetObjectDefaults_PodTemplate(in *v1.PodTemplate) {
if a.VolumeSource.ScaleIO != nil {
SetDefaults_ScaleIOVolumeSource(a.VolumeSource.ScaleIO)
}
if a.VolumeSource.Ephemeral != nil {
if a.VolumeSource.Ephemeral.VolumeClaimTemplate != nil {
SetDefaults_PersistentVolumeClaimSpec(&a.VolumeSource.Ephemeral.VolumeClaimTemplate.Spec)
SetDefaults_ResourceList(&a.VolumeSource.Ephemeral.VolumeClaimTemplate.Spec.Resources.Limits)
SetDefaults_ResourceList(&a.VolumeSource.Ephemeral.VolumeClaimTemplate.Spec.Resources.Requests)
}
}
}
for i := range in.Template.Spec.InitContainers {
a := &in.Template.Spec.InitContainers[i]
SetDefaults_Container(a)
for j := range a.Ports {
b := &a.Ports[j]
SetDefaults_ContainerPort(b)
if reflect.ValueOf(b.Protocol).IsZero() {
b.Protocol = "TCP"
}
}
for j := range a.Env {
b := &a.Env[j]
@ -532,7 +561,9 @@ func SetObjectDefaults_PodTemplate(in *v1.PodTemplate) {
SetDefaults_Container(a)
for j := range a.Ports {
b := &a.Ports[j]
SetDefaults_ContainerPort(b)
if reflect.ValueOf(b.Protocol).IsZero() {
b.Protocol = "TCP"
}
}
for j := range a.Env {
b := &a.Env[j]
@ -577,9 +608,12 @@ func SetObjectDefaults_PodTemplate(in *v1.PodTemplate) {
}
for i := range in.Template.Spec.EphemeralContainers {
a := &in.Template.Spec.EphemeralContainers[i]
SetDefaults_EphemeralContainer(a)
for j := range a.EphemeralContainerCommon.Ports {
b := &a.EphemeralContainerCommon.Ports[j]
SetDefaults_ContainerPort(b)
if reflect.ValueOf(b.Protocol).IsZero() {
b.Protocol = "TCP"
}
}
for j := range a.EphemeralContainerCommon.Env {
b := &a.EphemeralContainerCommon.Env[j]
@ -686,13 +720,22 @@ func SetObjectDefaults_ReplicationController(in *v1.ReplicationController) {
if a.VolumeSource.ScaleIO != nil {
SetDefaults_ScaleIOVolumeSource(a.VolumeSource.ScaleIO)
}
if a.VolumeSource.Ephemeral != nil {
if a.VolumeSource.Ephemeral.VolumeClaimTemplate != nil {
SetDefaults_PersistentVolumeClaimSpec(&a.VolumeSource.Ephemeral.VolumeClaimTemplate.Spec)
SetDefaults_ResourceList(&a.VolumeSource.Ephemeral.VolumeClaimTemplate.Spec.Resources.Limits)
SetDefaults_ResourceList(&a.VolumeSource.Ephemeral.VolumeClaimTemplate.Spec.Resources.Requests)
}
}
}
for i := range in.Spec.Template.Spec.InitContainers {
a := &in.Spec.Template.Spec.InitContainers[i]
SetDefaults_Container(a)
for j := range a.Ports {
b := &a.Ports[j]
SetDefaults_ContainerPort(b)
if reflect.ValueOf(b.Protocol).IsZero() {
b.Protocol = "TCP"
}
}
for j := range a.Env {
b := &a.Env[j]
@ -740,7 +783,9 @@ func SetObjectDefaults_ReplicationController(in *v1.ReplicationController) {
SetDefaults_Container(a)
for j := range a.Ports {
b := &a.Ports[j]
SetDefaults_ContainerPort(b)
if reflect.ValueOf(b.Protocol).IsZero() {
b.Protocol = "TCP"
}
}
for j := range a.Env {
b := &a.Env[j]
@ -785,9 +830,12 @@ func SetObjectDefaults_ReplicationController(in *v1.ReplicationController) {
}
for i := range in.Spec.Template.Spec.EphemeralContainers {
a := &in.Spec.Template.Spec.EphemeralContainers[i]
SetDefaults_EphemeralContainer(a)
for j := range a.EphemeralContainerCommon.Ports {
b := &a.EphemeralContainerCommon.Ports[j]
SetDefaults_ContainerPort(b)
if reflect.ValueOf(b.Protocol).IsZero() {
b.Protocol = "TCP"
}
}
for j := range a.EphemeralContainerCommon.Env {
b := &a.EphemeralContainerCommon.Env[j]

View File

@ -22,24 +22,26 @@ go_library(
"//pkg/apis/core/v1:go_default_library",
"//pkg/apis/core/v1/helper:go_default_library",
"//pkg/capabilities:go_default_library",
"//pkg/cluster/ports:go_default_library",
"//pkg/features:go_default_library",
"//pkg/fieldpath:go_default_library",
"//pkg/master/ports:go_default_library",
"//pkg/security/apparmor:go_default_library",
"//staging/src/k8s.io/api/core/v1:go_default_library",
"//staging/src/k8s.io/api/events/v1beta1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/api/equality:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/api/resource:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/api/validation:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/apis/meta/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/apis/meta/v1/validation:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/labels:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/runtime/schema:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/diff:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/intstr:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/sets:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/validation:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/validation/field:go_default_library",
"//staging/src/k8s.io/apiserver/pkg/util/feature:go_default_library",
"//vendor/k8s.io/klog:go_default_library",
"//vendor/k8s.io/klog/v2:go_default_library",
"//vendor/k8s.io/utils/net:go_default_library",
],
)
@ -56,16 +58,18 @@ go_test(
"//pkg/apis/core:go_default_library",
"//pkg/capabilities:go_default_library",
"//pkg/features:go_default_library",
"//pkg/security/apparmor:go_default_library",
"//staging/src/k8s.io/api/core/v1:go_default_library",
"//staging/src/k8s.io/api/events/v1:go_default_library",
"//staging/src/k8s.io/api/events/v1beta1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/api/resource:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/apis/meta/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/diff:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/intstr:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/validation:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/validation/field:go_default_library",
"//staging/src/k8s.io/apiserver/pkg/util/feature:go_default_library",
"//staging/src/k8s.io/component-base/featuregate/testing:go_default_library",
"//vendor/github.com/stretchr/testify/assert:go_default_library",
"//vendor/github.com/stretchr/testify/require:go_default_library",
"//vendor/k8s.io/utils/pointer:go_default_library",
],
)

View File

@ -21,12 +21,10 @@ reviewers:
- sttts
- quinton-hoole
- dchen1107
- zmerlynn
- janetkuo
- justinsb
- pwittrock
- tallclair
- eparis
- soltysh
- piosz
- jsafrane

View File

@ -26,123 +26,36 @@ import (
// ValidateConditionalService validates conditionally valid fields.
func ValidateConditionalService(service, oldService *api.Service) field.ErrorList {
var errs field.ErrorList
// If the SCTPSupport feature is disabled, and the old object isn't using the SCTP feature, prevent the new object from using it
if !utilfeature.DefaultFeatureGate.Enabled(features.SCTPSupport) && len(serviceSCTPFields(oldService)) == 0 {
for _, f := range serviceSCTPFields(service) {
errs = append(errs, field.NotSupported(f, api.ProtocolSCTP, []string{string(api.ProtocolTCP), string(api.ProtocolUDP)}))
}
}
errs = append(errs, validateMixedProtocolLBService(service, oldService)...)
return errs
}
func serviceSCTPFields(service *api.Service) []*field.Path {
// validateMixedProtocolLBService checks if the old Service has type=LoadBalancer and whether the Service has different Protocols
// on its ports. If the MixedProtocolLBService feature flag is disabled the usage of different Protocols in the new Service is
// valid only if the old Service has different Protocols, too.
func validateMixedProtocolLBService(service, oldService *api.Service) (errs field.ErrorList) {
if service.Spec.Type != api.ServiceTypeLoadBalancer {
return
}
if utilfeature.DefaultFeatureGate.Enabled(features.MixedProtocolLBService) {
return
}
if serviceHasMixedProtocols(service) && !serviceHasMixedProtocols(oldService) {
errs = append(errs, field.Invalid(field.NewPath("spec", "ports"), service.Spec.Ports, "may not contain more than 1 protocol when type is 'LoadBalancer'"))
}
return
}
func serviceHasMixedProtocols(service *api.Service) bool {
if service == nil {
return nil
return false
}
fields := []*field.Path{}
for pIndex, p := range service.Spec.Ports {
if p.Protocol == api.ProtocolSCTP {
fields = append(fields, field.NewPath("spec.ports").Index(pIndex).Child("protocol"))
}
protos := map[string]bool{}
for _, port := range service.Spec.Ports {
protos[string(port.Protocol)] = true
}
return fields
}
// ValidateConditionalEndpoints validates conditionally valid fields.
func ValidateConditionalEndpoints(endpoints, oldEndpoints *api.Endpoints) field.ErrorList {
var errs field.ErrorList
// If the SCTPSupport feature is disabled, and the old object isn't using the SCTP feature, prevent the new object from using it
if !utilfeature.DefaultFeatureGate.Enabled(features.SCTPSupport) && len(endpointsSCTPFields(oldEndpoints)) == 0 {
for _, f := range endpointsSCTPFields(endpoints) {
errs = append(errs, field.NotSupported(f, api.ProtocolSCTP, []string{string(api.ProtocolTCP), string(api.ProtocolUDP)}))
}
}
return errs
}
func endpointsSCTPFields(endpoints *api.Endpoints) []*field.Path {
if endpoints == nil {
return nil
}
fields := []*field.Path{}
for sIndex, s := range endpoints.Subsets {
for pIndex, p := range s.Ports {
if p.Protocol == api.ProtocolSCTP {
fields = append(fields, field.NewPath("subsets").Index(sIndex).Child("ports").Index(pIndex).Child("protocol"))
}
}
}
return fields
}
// ValidateConditionalPodTemplate validates conditionally valid fields.
// This should be called from Validate/ValidateUpdate for all resources containing a PodTemplateSpec
func ValidateConditionalPodTemplate(podTemplate, oldPodTemplate *api.PodTemplateSpec, fldPath *field.Path) field.ErrorList {
var (
podSpec *api.PodSpec
oldPodSpec *api.PodSpec
)
if podTemplate != nil {
podSpec = &podTemplate.Spec
}
if oldPodTemplate != nil {
oldPodSpec = &oldPodTemplate.Spec
}
return validateConditionalPodSpec(podSpec, oldPodSpec, fldPath.Child("spec"))
}
// ValidateConditionalPod validates conditionally valid fields.
// This should be called from Validate/ValidateUpdate for all resources containing a Pod
func ValidateConditionalPod(pod, oldPod *api.Pod, fldPath *field.Path) field.ErrorList {
var (
podSpec *api.PodSpec
oldPodSpec *api.PodSpec
)
if pod != nil {
podSpec = &pod.Spec
}
if oldPod != nil {
oldPodSpec = &oldPod.Spec
}
return validateConditionalPodSpec(podSpec, oldPodSpec, fldPath.Child("spec"))
}
func validateConditionalPodSpec(podSpec, oldPodSpec *api.PodSpec, fldPath *field.Path) field.ErrorList {
// Always make sure we have a non-nil current pod spec
if podSpec == nil {
podSpec = &api.PodSpec{}
}
errs := field.ErrorList{}
// If the SCTPSupport feature is disabled, and the old object isn't using the SCTP feature, prevent the new object from using it
if !utilfeature.DefaultFeatureGate.Enabled(features.SCTPSupport) && len(podSCTPFields(oldPodSpec, nil)) == 0 {
for _, f := range podSCTPFields(podSpec, fldPath) {
errs = append(errs, field.NotSupported(f, api.ProtocolSCTP, []string{string(api.ProtocolTCP), string(api.ProtocolUDP)}))
}
}
return errs
}
func podSCTPFields(podSpec *api.PodSpec, fldPath *field.Path) []*field.Path {
if podSpec == nil {
return nil
}
fields := []*field.Path{}
for cIndex, c := range podSpec.InitContainers {
for pIndex, p := range c.Ports {
if p.Protocol == api.ProtocolSCTP {
fields = append(fields, fldPath.Child("initContainers").Index(cIndex).Child("ports").Index(pIndex).Child("protocol"))
}
}
}
for cIndex, c := range podSpec.Containers {
for pIndex, p := range c.Ports {
if p.Protocol == api.ProtocolSCTP {
fields = append(fields, fldPath.Child("containers").Index(cIndex).Child("ports").Index(pIndex).Child("protocol"))
}
}
}
return fields
return len(protos) > 1
}

View File

@ -18,9 +18,14 @@ package validation
import (
"fmt"
"reflect"
"time"
"k8s.io/api/core/v1"
eventsv1beta1 "k8s.io/api/events/v1beta1"
apimachineryvalidation "k8s.io/apimachinery/pkg/api/validation"
metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
"k8s.io/apimachinery/pkg/runtime/schema"
"k8s.io/apimachinery/pkg/util/validation"
"k8s.io/apimachinery/pkg/util/validation/field"
"k8s.io/kubernetes/pkg/apis/core"
@ -33,8 +38,88 @@ const (
NoteLengthLimit = 1024
)
// ValidateEvent makes sure that the event makes sense.
func ValidateEvent(event *core.Event) field.ErrorList {
func ValidateEventCreate(event *core.Event, requestVersion schema.GroupVersion) field.ErrorList {
// Make sure events always pass legacy validation.
allErrs := legacyValidateEvent(event)
if requestVersion == v1.SchemeGroupVersion || requestVersion == eventsv1beta1.SchemeGroupVersion {
// No further validation for backwards compatibility.
return allErrs
}
// Strict validation applies to creation via events.k8s.io/v1 API and newer.
allErrs = append(allErrs, ValidateObjectMeta(&event.ObjectMeta, true, apimachineryvalidation.NameIsDNSSubdomain, field.NewPath("metadata"))...)
allErrs = append(allErrs, validateV1EventSeries(event)...)
zeroTime := time.Time{}
if event.EventTime.Time == zeroTime {
allErrs = append(allErrs, field.Required(field.NewPath("eventTime"), ""))
}
if event.Type != v1.EventTypeNormal && event.Type != v1.EventTypeWarning {
allErrs = append(allErrs, field.Invalid(field.NewPath("type"), "", fmt.Sprintf("has invalid value: %v", event.Type)))
}
if event.FirstTimestamp.Time != zeroTime {
allErrs = append(allErrs, field.Invalid(field.NewPath("firstTimestamp"), "", "needs to be unset"))
}
if event.LastTimestamp.Time != zeroTime {
allErrs = append(allErrs, field.Invalid(field.NewPath("lastTimestamp"), "", "needs to be unset"))
}
if event.Count != 0 {
allErrs = append(allErrs, field.Invalid(field.NewPath("count"), "", "needs to be unset"))
}
if event.Source.Component != "" || event.Source.Host != "" {
allErrs = append(allErrs, field.Invalid(field.NewPath("source"), "", "needs to be unset"))
}
return allErrs
}
func ValidateEventUpdate(newEvent, oldEvent *core.Event, requestVersion schema.GroupVersion) field.ErrorList {
// Make sure the new event always passes legacy validation.
allErrs := legacyValidateEvent(newEvent)
if requestVersion == v1.SchemeGroupVersion || requestVersion == eventsv1beta1.SchemeGroupVersion {
// No further validation for backwards compatibility.
return allErrs
}
// Strict validation applies to update via events.k8s.io/v1 API and newer.
allErrs = append(allErrs, ValidateObjectMetaUpdate(&newEvent.ObjectMeta, &oldEvent.ObjectMeta, field.NewPath("metadata"))...)
// if the series was modified, validate the new data
if !reflect.DeepEqual(newEvent.Series, oldEvent.Series) {
allErrs = append(allErrs, validateV1EventSeries(newEvent)...)
}
allErrs = append(allErrs, ValidateImmutableField(newEvent.InvolvedObject, oldEvent.InvolvedObject, field.NewPath("involvedObject"))...)
allErrs = append(allErrs, ValidateImmutableField(newEvent.Reason, oldEvent.Reason, field.NewPath("reason"))...)
allErrs = append(allErrs, ValidateImmutableField(newEvent.Message, oldEvent.Message, field.NewPath("message"))...)
allErrs = append(allErrs, ValidateImmutableField(newEvent.Source, oldEvent.Source, field.NewPath("source"))...)
allErrs = append(allErrs, ValidateImmutableField(newEvent.FirstTimestamp, oldEvent.FirstTimestamp, field.NewPath("firstTimestamp"))...)
allErrs = append(allErrs, ValidateImmutableField(newEvent.LastTimestamp, oldEvent.LastTimestamp, field.NewPath("lastTimestamp"))...)
allErrs = append(allErrs, ValidateImmutableField(newEvent.Count, oldEvent.Count, field.NewPath("count"))...)
allErrs = append(allErrs, ValidateImmutableField(newEvent.Reason, oldEvent.Reason, field.NewPath("reason"))...)
allErrs = append(allErrs, ValidateImmutableField(newEvent.Type, oldEvent.Type, field.NewPath("type"))...)
allErrs = append(allErrs, ValidateImmutableField(newEvent.EventTime, oldEvent.EventTime, field.NewPath("eventTime"))...)
allErrs = append(allErrs, ValidateImmutableField(newEvent.Action, oldEvent.Action, field.NewPath("action"))...)
allErrs = append(allErrs, ValidateImmutableField(newEvent.Related, oldEvent.Related, field.NewPath("related"))...)
allErrs = append(allErrs, ValidateImmutableField(newEvent.ReportingController, oldEvent.ReportingController, field.NewPath("reportingController"))...)
allErrs = append(allErrs, ValidateImmutableField(newEvent.ReportingInstance, oldEvent.ReportingInstance, field.NewPath("reportingInstance"))...)
return allErrs
}
func validateV1EventSeries(event *core.Event) field.ErrorList {
allErrs := field.ErrorList{}
zeroTime := time.Time{}
if event.Series != nil {
if event.Series.Count < 2 {
allErrs = append(allErrs, field.Invalid(field.NewPath("series.count"), "", fmt.Sprintf("should be at least 2")))
}
if event.Series.LastObservedTime.Time == zeroTime {
allErrs = append(allErrs, field.Required(field.NewPath("series.lastObservedTime"), ""))
}
}
return allErrs
}
// legacyValidateEvent makes sure that the event makes sense.
func legacyValidateEvent(event *core.Event) field.ErrorList {
allErrs := field.ErrorList{}
// Because go
zeroTime := time.Time{}

File diff suppressed because it is too large Load Diff

View File

@ -1433,6 +1433,27 @@ func (in *EphemeralContainers) DeepCopyObject() runtime.Object {
return nil
}
// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
func (in *EphemeralVolumeSource) DeepCopyInto(out *EphemeralVolumeSource) {
*out = *in
if in.VolumeClaimTemplate != nil {
in, out := &in.VolumeClaimTemplate, &out.VolumeClaimTemplate
*out = new(PersistentVolumeClaimTemplate)
(*in).DeepCopyInto(*out)
}
return
}
// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new EphemeralVolumeSource.
func (in *EphemeralVolumeSource) DeepCopy() *EphemeralVolumeSource {
if in == nil {
return nil
}
out := new(EphemeralVolumeSource)
in.DeepCopyInto(out)
return out
}
// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
func (in *Event) DeepCopyInto(out *Event) {
*out = *in
@ -2125,6 +2146,13 @@ func (in *List) DeepCopyObject() runtime.Object {
// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
func (in *LoadBalancerIngress) DeepCopyInto(out *LoadBalancerIngress) {
*out = *in
if in.Ports != nil {
in, out := &in.Ports, &out.Ports
*out = make([]PortStatus, len(*in))
for i := range *in {
(*in)[i].DeepCopyInto(&(*out)[i])
}
}
return
}
@ -2144,7 +2172,9 @@ func (in *LoadBalancerStatus) DeepCopyInto(out *LoadBalancerStatus) {
if in.Ingress != nil {
in, out := &in.Ingress, &out.Ingress
*out = make([]LoadBalancerIngress, len(*in))
copy(*out, *in)
for i := range *in {
(*in)[i].DeepCopyInto(&(*out)[i])
}
}
return
}
@ -2987,6 +3017,24 @@ func (in *PersistentVolumeClaimStatus) DeepCopy() *PersistentVolumeClaimStatus {
return out
}
// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
func (in *PersistentVolumeClaimTemplate) DeepCopyInto(out *PersistentVolumeClaimTemplate) {
*out = *in
in.ObjectMeta.DeepCopyInto(&out.ObjectMeta)
in.Spec.DeepCopyInto(&out.Spec)
return
}
// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new PersistentVolumeClaimTemplate.
func (in *PersistentVolumeClaimTemplate) DeepCopy() *PersistentVolumeClaimTemplate {
if in == nil {
return nil
}
out := new(PersistentVolumeClaimTemplate)
in.DeepCopyInto(out)
return out
}
// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
func (in *PersistentVolumeClaimVolumeSource) DeepCopyInto(out *PersistentVolumeClaimVolumeSource) {
*out = *in
@ -3701,6 +3749,11 @@ func (in *PodSecurityContext) DeepCopyInto(out *PodSecurityContext) {
*out = make([]Sysctl, len(*in))
copy(*out, *in)
}
if in.SeccompProfile != nil {
in, out := &in.SeccompProfile, &out.SeccompProfile
*out = new(SeccompProfile)
(*in).DeepCopyInto(*out)
}
return
}
@ -3798,6 +3851,11 @@ func (in *PodSpec) DeepCopyInto(out *PodSpec) {
*out = make([]LocalObjectReference, len(*in))
copy(*out, *in)
}
if in.SetHostnameAsFQDN != nil {
in, out := &in.SetHostnameAsFQDN, &out.SetHostnameAsFQDN
*out = new(bool)
**out = **in
}
if in.Affinity != nil {
in, out := &in.Affinity, &out.Affinity
*out = new(Affinity)
@ -4032,6 +4090,27 @@ func (in *PodTemplateSpec) DeepCopy() *PodTemplateSpec {
return out
}
// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
func (in *PortStatus) DeepCopyInto(out *PortStatus) {
*out = *in
if in.Error != nil {
in, out := &in.Error, &out.Error
*out = new(string)
**out = **in
}
return
}
// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new PortStatus.
func (in *PortStatus) DeepCopy() *PortStatus {
if in == nil {
return nil
}
out := new(PortStatus)
in.DeepCopyInto(out)
return out
}
// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
func (in *PortworxVolumeSource) DeepCopyInto(out *PortworxVolumeSource) {
*out = *in
@ -4672,6 +4751,27 @@ func (in *ScopedResourceSelectorRequirement) DeepCopy() *ScopedResourceSelectorR
return out
}
// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
func (in *SeccompProfile) DeepCopyInto(out *SeccompProfile) {
*out = *in
if in.LocalhostProfile != nil {
in, out := &in.LocalhostProfile, &out.LocalhostProfile
*out = new(string)
**out = **in
}
return
}
// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new SeccompProfile.
func (in *SeccompProfile) DeepCopy() *SeccompProfile {
if in == nil {
return nil
}
out := new(SeccompProfile)
in.DeepCopyInto(out)
return out
}
// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
func (in *Secret) DeepCopyInto(out *Secret) {
*out = *in
@ -4926,6 +5026,11 @@ func (in *SecurityContext) DeepCopyInto(out *SecurityContext) {
*out = new(ProcMountType)
**out = **in
}
if in.SeccompProfile != nil {
in, out := &in.SeccompProfile, &out.SeccompProfile
*out = new(SeccompProfile)
(*in).DeepCopyInto(*out)
}
return
}
@ -5180,6 +5285,21 @@ func (in *ServiceSpec) DeepCopyInto(out *ServiceSpec) {
(*out)[key] = val
}
}
if in.ClusterIPs != nil {
in, out := &in.ClusterIPs, &out.ClusterIPs
*out = make([]string, len(*in))
copy(*out, *in)
}
if in.IPFamilies != nil {
in, out := &in.IPFamilies, &out.IPFamilies
*out = make([]IPFamily, len(*in))
copy(*out, *in)
}
if in.IPFamilyPolicy != nil {
in, out := &in.IPFamilyPolicy, &out.IPFamilyPolicy
*out = new(IPFamilyPolicyType)
**out = **in
}
if in.ExternalIPs != nil {
in, out := &in.ExternalIPs, &out.ExternalIPs
*out = make([]string, len(*in))
@ -5195,16 +5315,16 @@ func (in *ServiceSpec) DeepCopyInto(out *ServiceSpec) {
*out = make([]string, len(*in))
copy(*out, *in)
}
if in.IPFamily != nil {
in, out := &in.IPFamily, &out.IPFamily
*out = new(IPFamily)
**out = **in
}
if in.TopologyKeys != nil {
in, out := &in.TopologyKeys, &out.TopologyKeys
*out = make([]string, len(*in))
copy(*out, *in)
}
if in.AllocateLoadBalancerNodePorts != nil {
in, out := &in.AllocateLoadBalancerNodePorts, &out.AllocateLoadBalancerNodePorts
*out = new(bool)
**out = **in
}
return
}
@ -5222,6 +5342,13 @@ func (in *ServiceSpec) DeepCopy() *ServiceSpec {
func (in *ServiceStatus) DeepCopyInto(out *ServiceStatus) {
*out = *in
in.LoadBalancer.DeepCopyInto(&out.LoadBalancer)
if in.Conditions != nil {
in, out := &in.Conditions, &out.Conditions
*out = make([]v1.Condition, len(*in))
for i := range *in {
(*in)[i].DeepCopyInto(&(*out)[i])
}
}
return
}
@ -5712,6 +5839,11 @@ func (in *VolumeSource) DeepCopyInto(out *VolumeSource) {
*out = new(CSIVolumeSource)
(*in).DeepCopyInto(*out)
}
if in.Ephemeral != nil {
in, out := &in.Ephemeral, &out.Ephemeral
*out = new(EphemeralVolumeSource)
(*in).DeepCopyInto(*out)
}
return
}

View File

@ -30,10 +30,8 @@ reviewers:
- mml
- resouer
- mbohlool
- david-mcmahon
- therc
- pweil-
- mqliang
- lukaszo
- jianhuiz
labels:

View File

@ -158,7 +158,7 @@ type IPBlock struct {
Except []string
}
// NetworkPolicyPeer describes a peer to allow traffic from.
// NetworkPolicyPeer describes a peer to allow traffic to/from.
type NetworkPolicyPeer struct {
// This is a label selector which selects Pods. This field follows standard label
// selector semantics; if present but empty, it selects all pods.
@ -248,26 +248,27 @@ type IngressSpec struct {
// +optional
IngressClassName *string
// Backend is a default backend capable of servicing requests that don't
// match any rule. At least one of 'backend' or 'rules' must be specified.
// This field is optional to allow the loadbalancer controller or defaulting
// logic to specify a global default.
// DefaultBackend is the backend that should handle requests that don't
// match any rule. If Rules are not specified, DefaultBackend must be specified.
// If DefaultBackend is not set, the handling of requests that do not match any
// of the rules will be up to the Ingress controller.
// +optional
Backend *IngressBackend
DefaultBackend *IngressBackend
// TLS configuration. Currently the Ingress only supports a single TLS
// port, 443. If multiple members of this list specify different hosts, they
// will be multiplexed on the same port according to the hostname specified
// through the SNI TLS extension, if the ingress controller fulfilling the
// ingress supports SNI.
// +listType=atomic
// +optional
TLS []IngressTLS
// A list of host rules used to configure the Ingress. If unspecified, or
// no rule matches, all traffic is sent to the default backend.
// +listType=atomic
// +optional
Rules []IngressRule
// TODO: Add the ability to specify load-balancer IP through claims
}
// +k8s:deepcopy-gen:interfaces=k8s.io/apimachinery/pkg/runtime.Object
@ -314,7 +315,6 @@ type IngressClassList struct {
metav1.ListMeta
// Items is the list of IngressClasses.
// +listType=set
Items []IngressClass
}
@ -324,6 +324,7 @@ type IngressTLS struct {
// this list must match the name/s used in the tlsSecret. Defaults to the
// wildcard host setting for the loadbalancer controller fulfilling this
// Ingress, if left unspecified.
// +listType=atomic
// +optional
Hosts []string
// SecretName is the name of the secret used to terminate TLS traffic on
@ -404,6 +405,7 @@ type IngressRuleValue struct {
// or '#'.
type HTTPIngressRuleValue struct {
// A collection of paths that map requests to backends.
// +listType=atomic
Paths []HTTPIngressPath
// TODO: Consider adding fields for ingress-type specific global
// options usable by a loadbalancer, like http keep-alive.
@ -462,17 +464,39 @@ type HTTPIngressPath struct {
// IngressBackend describes all endpoints for a given service and port.
type IngressBackend struct {
// Specifies the name of the referenced service.
// Service references a Service as a Backend.
// This is a mutually exclusive setting with "Resource".
// +optional
ServiceName string
// Specifies the port of the referenced service.
// +optional
ServicePort intstr.IntOrString
Service *IngressServiceBackend
// Resource is an ObjectRef to another Kubernetes resource in the namespace
// of the Ingress object. If resource is specified, serviceName and servicePort
// must not be specified.
// of the Ingress object. If resource is specified, a service.Name and
// service.Port must not be specified.
// This is a mutually exclusive setting with "Service".
// +optional
Resource *api.TypedLocalObjectReference
}
// IngressServiceBackend references a Kubernetes Service as a Backend.
type IngressServiceBackend struct {
// Name is the referenced service. The service must exist in
// the same namespace as the Ingress object.
Name string
// Port of the referenced service. A port name or port number
// is required for a IngressServiceBackend.
Port ServiceBackendPort
}
// ServiceBackendPort is the service port being referenced.
type ServiceBackendPort struct {
// Name is the name of the port on the Service.
// This is a mutually exclusive setting with "Number".
// +optional
Name string
// Number is the numerical port number (e.g. 80) on the Service.
// This is a mutually exclusive setting with "Name".
// +optional
Number int32
}

View File

@ -124,7 +124,11 @@ func (in *Ingress) DeepCopyObject() runtime.Object {
// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
func (in *IngressBackend) DeepCopyInto(out *IngressBackend) {
*out = *in
out.ServicePort = in.ServicePort
if in.Service != nil {
in, out := &in.Service, &out.Service
*out = new(IngressServiceBackend)
**out = **in
}
if in.Resource != nil {
in, out := &in.Resource, &out.Resource
*out = new(core.TypedLocalObjectReference)
@ -295,6 +299,23 @@ func (in *IngressRuleValue) DeepCopy() *IngressRuleValue {
return out
}
// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
func (in *IngressServiceBackend) DeepCopyInto(out *IngressServiceBackend) {
*out = *in
out.Port = in.Port
return
}
// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new IngressServiceBackend.
func (in *IngressServiceBackend) DeepCopy() *IngressServiceBackend {
if in == nil {
return nil
}
out := new(IngressServiceBackend)
in.DeepCopyInto(out)
return out
}
// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
func (in *IngressSpec) DeepCopyInto(out *IngressSpec) {
*out = *in
@ -303,8 +324,8 @@ func (in *IngressSpec) DeepCopyInto(out *IngressSpec) {
*out = new(string)
**out = **in
}
if in.Backend != nil {
in, out := &in.Backend, &out.Backend
if in.DefaultBackend != nil {
in, out := &in.DefaultBackend, &out.DefaultBackend
*out = new(IngressBackend)
(*in).DeepCopyInto(*out)
}
@ -585,3 +606,19 @@ func (in *NetworkPolicySpec) DeepCopy() *NetworkPolicySpec {
in.DeepCopyInto(out)
return out
}
// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
func (in *ServiceBackendPort) DeepCopyInto(out *ServiceBackendPort) {
*out = *in
return
}
// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new ServiceBackendPort.
func (in *ServiceBackendPort) DeepCopy() *ServiceBackendPort {
if in == nil {
return nil
}
out := new(ServiceBackendPort)
in.DeepCopyInto(out)
return out
}

View File

@ -157,7 +157,7 @@ type PodSecurityPolicySpec struct {
// To allow all capabilities you may use '*'.
// +optional
AllowedCapabilities []api.Capability
// Volumes is a white list of allowed volume plugins. Empty indicates that
// Volumes is an allowlist of volume plugins. Empty indicates that
// no volumes may be used. To allow all volumes you may use '*'.
// +optional
Volumes []FSType
@ -200,23 +200,23 @@ type PodSecurityPolicySpec struct {
// privilege escalation. If unspecified, defaults to true.
// +optional
AllowPrivilegeEscalation bool
// AllowedHostPaths is a white list of allowed host paths. Empty indicates that all host paths may be used.
// AllowedHostPaths is an allowlist of host paths. Empty indicates that all host paths may be used.
// +optional
AllowedHostPaths []AllowedHostPath
// AllowedFlexVolumes is a whitelist of allowed Flexvolumes. Empty or nil indicates that all
// AllowedFlexVolumes is an allowlist of Flexvolumes. Empty or nil indicates that all
// Flexvolumes may be used. This parameter is effective only when the usage of the Flexvolumes
// is allowed in the "Volumes" field.
// +optional
AllowedFlexVolumes []AllowedFlexVolume
// AllowedCSIDrivers is a whitelist of inline CSI drivers that must be explicitly set to be embedded within a pod spec.
// AllowedCSIDrivers is an allowlist of inline CSI drivers that must be explicitly set to be embedded within a pod spec.
// An empty value indicates that any CSI driver can be used for inline ephemeral volumes.
// This is an alpha field, and is only honored if the API server enables the CSIInlineVolume feature gate.
// This is a beta field, and is only honored if the API server enables the CSIInlineVolume feature gate.
// +optional
AllowedCSIDrivers []AllowedCSIDriver
// AllowedUnsafeSysctls is a list of explicitly allowed unsafe sysctls, defaults to none.
// Each entry is either a plain sysctl name or ends in "*" in which case it is considered
// as a prefix of allowed sysctls. Single * means all unsafe sysctls are allowed.
// Kubelet has to whitelist all allowed unsafe sysctls explicitly to avoid rejection.
// Kubelet has to allowlist all unsafe sysctls explicitly to avoid rejection.
//
// Examples:
// e.g. "foo/*" allows "foo/bar", "foo/baz", etc.
@ -232,7 +232,7 @@ type PodSecurityPolicySpec struct {
// e.g. "foo.*" forbids "foo.bar", "foo.baz", etc.
// +optional
ForbiddenSysctls []string
// AllowedProcMountTypes is a whitelist of allowed ProcMountTypes.
// AllowedProcMountTypes is an allowlist of ProcMountTypes.
// Empty or nil indicates that only the DefaultProcMountType may be used.
// +optional
AllowedProcMountTypes []api.ProcMountType
@ -449,7 +449,7 @@ const (
// RuntimeClassStrategyOptions define the strategy that will dictate the allowable RuntimeClasses
// for a pod.
type RuntimeClassStrategyOptions struct {
// allowedRuntimeClassNames is a whitelist of RuntimeClass names that may be specified on a pod.
// allowedRuntimeClassNames is an allowlist of RuntimeClass names that may be specified on a pod.
// A value of "*" means that any RuntimeClass name is allowed, and must be the only item in the
// list. An empty list requires the RuntimeClassName field to be unset.
AllowedRuntimeClassNames []string

View File

@ -1,57 +0,0 @@
package(default_visibility = ["//visibility:public"])
load(
"@io_bazel_rules_go//go:def.bzl",
"go_library",
"go_test",
)
go_library(
name = "go_default_library",
srcs = ["validation.go"],
importpath = "k8s.io/kubernetes/pkg/apis/policy/validation",
deps = [
"//pkg/apis/apps/validation:go_default_library",
"//pkg/apis/core:go_default_library",
"//pkg/apis/core/validation:go_default_library",
"//pkg/apis/policy:go_default_library",
"//pkg/security/apparmor:go_default_library",
"//pkg/security/podsecuritypolicy/seccomp:go_default_library",
"//pkg/security/podsecuritypolicy/util:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/api/validation:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/apis/meta/v1/validation:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/sets:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/validation/field:go_default_library",
],
)
go_test(
name = "go_default_test",
srcs = ["validation_test.go"],
embed = [":go_default_library"],
deps = [
"//pkg/apis/core:go_default_library",
"//pkg/apis/policy:go_default_library",
"//pkg/security/apparmor:go_default_library",
"//pkg/security/podsecuritypolicy/seccomp:go_default_library",
"//pkg/security/podsecuritypolicy/util:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/apis/meta/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/intstr:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/validation/field:go_default_library",
"//vendor/github.com/stretchr/testify/assert:go_default_library",
"//vendor/k8s.io/utils/pointer:go_default_library",
],
)
filegroup(
name = "package-srcs",
srcs = glob(["**"]),
tags = ["automanaged"],
visibility = ["//visibility:private"],
)
filegroup(
name = "all-srcs",
srcs = [":package-srcs"],
tags = ["automanaged"],
)

View File

@ -1,530 +0,0 @@
/*
Copyright 2016 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 validation
import (
"fmt"
"path/filepath"
"regexp"
"strings"
apimachineryvalidation "k8s.io/apimachinery/pkg/api/validation"
unversionedvalidation "k8s.io/apimachinery/pkg/apis/meta/v1/validation"
"k8s.io/apimachinery/pkg/util/sets"
"k8s.io/apimachinery/pkg/util/validation/field"
appsvalidation "k8s.io/kubernetes/pkg/apis/apps/validation"
core "k8s.io/kubernetes/pkg/apis/core"
apivalidation "k8s.io/kubernetes/pkg/apis/core/validation"
"k8s.io/kubernetes/pkg/apis/policy"
"k8s.io/kubernetes/pkg/security/apparmor"
"k8s.io/kubernetes/pkg/security/podsecuritypolicy/seccomp"
psputil "k8s.io/kubernetes/pkg/security/podsecuritypolicy/util"
)
// ValidatePodDisruptionBudget validates a PodDisruptionBudget and returns an ErrorList
// with any errors.
func ValidatePodDisruptionBudget(pdb *policy.PodDisruptionBudget) field.ErrorList {
allErrs := ValidatePodDisruptionBudgetSpec(pdb.Spec, field.NewPath("spec"))
allErrs = append(allErrs, ValidatePodDisruptionBudgetStatus(pdb.Status, field.NewPath("status"))...)
return allErrs
}
// ValidatePodDisruptionBudgetSpec validates a PodDisruptionBudgetSpec and returns an ErrorList
// with any errors.
func ValidatePodDisruptionBudgetSpec(spec policy.PodDisruptionBudgetSpec, fldPath *field.Path) field.ErrorList {
allErrs := field.ErrorList{}
if spec.MinAvailable != nil && spec.MaxUnavailable != nil {
allErrs = append(allErrs, field.Invalid(fldPath, spec, "minAvailable and maxUnavailable cannot be both set"))
}
if spec.MinAvailable != nil {
allErrs = append(allErrs, appsvalidation.ValidatePositiveIntOrPercent(*spec.MinAvailable, fldPath.Child("minAvailable"))...)
allErrs = append(allErrs, appsvalidation.IsNotMoreThan100Percent(*spec.MinAvailable, fldPath.Child("minAvailable"))...)
}
if spec.MaxUnavailable != nil {
allErrs = append(allErrs, appsvalidation.ValidatePositiveIntOrPercent(*spec.MaxUnavailable, fldPath.Child("maxUnavailable"))...)
allErrs = append(allErrs, appsvalidation.IsNotMoreThan100Percent(*spec.MaxUnavailable, fldPath.Child("maxUnavailable"))...)
}
allErrs = append(allErrs, unversionedvalidation.ValidateLabelSelector(spec.Selector, fldPath.Child("selector"))...)
return allErrs
}
// ValidatePodDisruptionBudgetStatus validates a PodDisruptionBudgetStatus and returns an ErrorList
// with any errors.
func ValidatePodDisruptionBudgetStatus(status policy.PodDisruptionBudgetStatus, fldPath *field.Path) field.ErrorList {
allErrs := field.ErrorList{}
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(status.DisruptionsAllowed), fldPath.Child("disruptionsAllowed"))...)
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(status.CurrentHealthy), fldPath.Child("currentHealthy"))...)
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(status.DesiredHealthy), fldPath.Child("desiredHealthy"))...)
allErrs = append(allErrs, apivalidation.ValidateNonnegativeField(int64(status.ExpectedPods), fldPath.Child("expectedPods"))...)
return allErrs
}
// ValidatePodSecurityPolicyName can be used to check whether the given
// pod security policy name is valid.
// Prefix indicates this name will be used as part of generation, in which case
// trailing dashes are allowed.
var ValidatePodSecurityPolicyName = apimachineryvalidation.NameIsDNSSubdomain
// ValidatePodSecurityPolicy validates a PodSecurityPolicy and returns an ErrorList
// with any errors.
func ValidatePodSecurityPolicy(psp *policy.PodSecurityPolicy) field.ErrorList {
allErrs := field.ErrorList{}
allErrs = append(allErrs, apivalidation.ValidateObjectMeta(&psp.ObjectMeta, false, ValidatePodSecurityPolicyName, field.NewPath("metadata"))...)
allErrs = append(allErrs, ValidatePodSecurityPolicySpecificAnnotations(psp.Annotations, field.NewPath("metadata").Child("annotations"))...)
allErrs = append(allErrs, ValidatePodSecurityPolicySpec(&psp.Spec, field.NewPath("spec"))...)
return allErrs
}
// ValidatePodSecurityPolicySpec validates a PodSecurityPolicySpec and returns an ErrorList
// with any errors.
func ValidatePodSecurityPolicySpec(spec *policy.PodSecurityPolicySpec, fldPath *field.Path) field.ErrorList {
allErrs := field.ErrorList{}
allErrs = append(allErrs, validatePSPRunAsUser(fldPath.Child("runAsUser"), &spec.RunAsUser)...)
allErrs = append(allErrs, validatePSPRunAsGroup(fldPath.Child("runAsGroup"), spec.RunAsGroup)...)
allErrs = append(allErrs, validatePSPSELinux(fldPath.Child("seLinux"), &spec.SELinux)...)
allErrs = append(allErrs, validatePSPSupplementalGroup(fldPath.Child("supplementalGroups"), &spec.SupplementalGroups)...)
allErrs = append(allErrs, validatePSPFSGroup(fldPath.Child("fsGroup"), &spec.FSGroup)...)
allErrs = append(allErrs, validatePodSecurityPolicyVolumes(fldPath, spec.Volumes)...)
if len(spec.RequiredDropCapabilities) > 0 && hasCap(policy.AllowAllCapabilities, spec.AllowedCapabilities) {
allErrs = append(allErrs, field.Invalid(field.NewPath("requiredDropCapabilities"), spec.RequiredDropCapabilities,
"must be empty when all capabilities are allowed by a wildcard"))
}
allErrs = append(allErrs, validatePSPCapsAgainstDrops(spec.RequiredDropCapabilities, spec.DefaultAddCapabilities, field.NewPath("defaultAddCapabilities"))...)
allErrs = append(allErrs, validatePSPCapsAgainstDrops(spec.RequiredDropCapabilities, spec.AllowedCapabilities, field.NewPath("allowedCapabilities"))...)
allErrs = append(allErrs, validatePSPDefaultAllowPrivilegeEscalation(fldPath.Child("defaultAllowPrivilegeEscalation"), spec.DefaultAllowPrivilegeEscalation, spec.AllowPrivilegeEscalation)...)
allErrs = append(allErrs, validatePSPAllowedProcMountTypes(fldPath.Child("allowedProcMountTypes"), spec.AllowedProcMountTypes)...)
allErrs = append(allErrs, validatePSPAllowedHostPaths(fldPath.Child("allowedHostPaths"), spec.AllowedHostPaths)...)
allErrs = append(allErrs, validatePSPAllowedFlexVolumes(fldPath.Child("allowedFlexVolumes"), spec.AllowedFlexVolumes)...)
allErrs = append(allErrs, validatePSPAllowedCSIDrivers(fldPath.Child("allowedCSIDrivers"), spec.AllowedCSIDrivers)...)
allErrs = append(allErrs, validatePodSecurityPolicySysctls(fldPath.Child("allowedUnsafeSysctls"), spec.AllowedUnsafeSysctls)...)
allErrs = append(allErrs, validatePodSecurityPolicySysctls(fldPath.Child("forbiddenSysctls"), spec.ForbiddenSysctls)...)
allErrs = append(allErrs, validatePodSecurityPolicySysctlListsDoNotOverlap(fldPath.Child("allowedUnsafeSysctls"), fldPath.Child("forbiddenSysctls"), spec.AllowedUnsafeSysctls, spec.ForbiddenSysctls)...)
allErrs = append(allErrs, validateRuntimeClassStrategy(fldPath.Child("runtimeClass"), spec.RuntimeClass)...)
return allErrs
}
// ValidatePodSecurityPolicySpecificAnnotations validates annotations and returns an ErrorList
// with any errors.
func ValidatePodSecurityPolicySpecificAnnotations(annotations map[string]string, fldPath *field.Path) field.ErrorList {
allErrs := field.ErrorList{}
if p := annotations[apparmor.DefaultProfileAnnotationKey]; p != "" {
if err := apparmor.ValidateProfileFormat(p); err != nil {
allErrs = append(allErrs, field.Invalid(fldPath.Key(apparmor.DefaultProfileAnnotationKey), p, err.Error()))
}
}
if allowed := annotations[apparmor.AllowedProfilesAnnotationKey]; allowed != "" {
for _, p := range strings.Split(allowed, ",") {
if err := apparmor.ValidateProfileFormat(p); err != nil {
allErrs = append(allErrs, field.Invalid(fldPath.Key(apparmor.AllowedProfilesAnnotationKey), allowed, err.Error()))
}
}
}
if p := annotations[seccomp.DefaultProfileAnnotationKey]; p != "" {
allErrs = append(allErrs, apivalidation.ValidateSeccompProfile(p, fldPath.Key(seccomp.DefaultProfileAnnotationKey))...)
}
if allowed := annotations[seccomp.AllowedProfilesAnnotationKey]; allowed != "" {
for _, p := range strings.Split(allowed, ",") {
if p == seccomp.AllowAny {
continue
}
allErrs = append(allErrs, apivalidation.ValidateSeccompProfile(p, fldPath.Key(seccomp.AllowedProfilesAnnotationKey))...)
}
}
return allErrs
}
// validatePSPAllowedHostPaths makes sure all allowed host paths follow:
// 1. path prefix is required
// 2. path prefix does not have any element which is ".."
func validatePSPAllowedHostPaths(fldPath *field.Path, allowedHostPaths []policy.AllowedHostPath) field.ErrorList {
allErrs := field.ErrorList{}
for i, target := range allowedHostPaths {
if target.PathPrefix == "" {
allErrs = append(allErrs, field.Required(fldPath.Index(i), "is required"))
break
}
parts := strings.Split(filepath.ToSlash(target.PathPrefix), "/")
for _, item := range parts {
if item == ".." {
allErrs = append(allErrs, field.Invalid(fldPath.Index(i), target.PathPrefix, "must not contain '..'"))
break // even for `../../..`, one error is sufficient to make the point
}
}
}
return allErrs
}
func validatePSPAllowedFlexVolumes(fldPath *field.Path, flexVolumes []policy.AllowedFlexVolume) field.ErrorList {
allErrs := field.ErrorList{}
if len(flexVolumes) > 0 {
for idx, fv := range flexVolumes {
if len(fv.Driver) == 0 {
allErrs = append(allErrs, field.Required(fldPath.Child("allowedFlexVolumes").Index(idx).Child("driver"),
"must specify a driver"))
}
}
}
return allErrs
}
func validatePSPAllowedCSIDrivers(fldPath *field.Path, csiDrivers []policy.AllowedCSIDriver) field.ErrorList {
allErrs := field.ErrorList{}
if len(csiDrivers) > 0 {
for idx, csiDriver := range csiDrivers {
fieldPath := fldPath.Child("allowedCSIDriver").Index(idx).Child("name")
allErrs = append(allErrs, apivalidation.ValidateCSIDriverName(csiDriver.Name, fieldPath)...)
}
}
return allErrs
}
// validatePSPSELinux validates the SELinux fields of PodSecurityPolicy.
func validatePSPSELinux(fldPath *field.Path, seLinux *policy.SELinuxStrategyOptions) field.ErrorList {
allErrs := field.ErrorList{}
// ensure the selinux strategy has a valid rule
supportedSELinuxRules := sets.NewString(
string(policy.SELinuxStrategyMustRunAs),
string(policy.SELinuxStrategyRunAsAny),
)
if !supportedSELinuxRules.Has(string(seLinux.Rule)) {
allErrs = append(allErrs, field.NotSupported(fldPath.Child("rule"), seLinux.Rule, supportedSELinuxRules.List()))
}
return allErrs
}
// validatePSPRunAsUser validates the RunAsUser fields of PodSecurityPolicy.
func validatePSPRunAsUser(fldPath *field.Path, runAsUser *policy.RunAsUserStrategyOptions) field.ErrorList {
allErrs := field.ErrorList{}
// ensure the user strategy has a valid rule
supportedRunAsUserRules := sets.NewString(
string(policy.RunAsUserStrategyMustRunAs),
string(policy.RunAsUserStrategyMustRunAsNonRoot),
string(policy.RunAsUserStrategyRunAsAny),
)
if !supportedRunAsUserRules.Has(string(runAsUser.Rule)) {
allErrs = append(allErrs, field.NotSupported(fldPath.Child("rule"), runAsUser.Rule, supportedRunAsUserRules.List()))
}
// validate range settings
for idx, rng := range runAsUser.Ranges {
allErrs = append(allErrs, validateUserIDRange(fldPath.Child("ranges").Index(idx), rng)...)
}
return allErrs
}
// validatePSPRunAsGroup validates the RunAsGroup fields of PodSecurityPolicy.
func validatePSPRunAsGroup(fldPath *field.Path, runAsGroup *policy.RunAsGroupStrategyOptions) field.ErrorList {
var allErrs field.ErrorList
if runAsGroup == nil {
return allErrs
}
switch runAsGroup.Rule {
case policy.RunAsGroupStrategyRunAsAny:
if len(runAsGroup.Ranges) != 0 {
allErrs = append(allErrs, field.Invalid(fldPath.Child("ranges"), runAsGroup.Ranges, "Ranges must be empty"))
}
case policy.RunAsGroupStrategyMustRunAs, policy.RunAsGroupStrategyMayRunAs:
if len(runAsGroup.Ranges) == 0 {
allErrs = append(allErrs, field.Invalid(fldPath.Child("ranges"), runAsGroup.Ranges, "must provide at least one range"))
}
// validate range settings
for idx, rng := range runAsGroup.Ranges {
allErrs = append(allErrs, validateGroupIDRange(fldPath.Child("ranges").Index(idx), rng)...)
}
default:
supportedRunAsGroupRules := []string{
string(policy.RunAsGroupStrategyMustRunAs),
string(policy.RunAsGroupStrategyRunAsAny),
string(policy.RunAsGroupStrategyMayRunAs),
}
allErrs = append(allErrs, field.NotSupported(fldPath.Child("rule"), runAsGroup.Rule, supportedRunAsGroupRules))
}
return allErrs
}
// validatePSPFSGroup validates the FSGroupStrategyOptions fields of the PodSecurityPolicy.
func validatePSPFSGroup(fldPath *field.Path, groupOptions *policy.FSGroupStrategyOptions) field.ErrorList {
allErrs := field.ErrorList{}
supportedRules := sets.NewString(
string(policy.FSGroupStrategyMustRunAs),
string(policy.FSGroupStrategyMayRunAs),
string(policy.FSGroupStrategyRunAsAny),
)
if !supportedRules.Has(string(groupOptions.Rule)) {
allErrs = append(allErrs, field.NotSupported(fldPath.Child("rule"), groupOptions.Rule, supportedRules.List()))
}
for idx, rng := range groupOptions.Ranges {
allErrs = append(allErrs, validateGroupIDRange(fldPath.Child("ranges").Index(idx), rng)...)
}
return allErrs
}
// validatePSPSupplementalGroup validates the SupplementalGroupsStrategyOptions fields of the PodSecurityPolicy.
func validatePSPSupplementalGroup(fldPath *field.Path, groupOptions *policy.SupplementalGroupsStrategyOptions) field.ErrorList {
allErrs := field.ErrorList{}
supportedRules := sets.NewString(
string(policy.SupplementalGroupsStrategyRunAsAny),
string(policy.SupplementalGroupsStrategyMayRunAs),
string(policy.SupplementalGroupsStrategyMustRunAs),
)
if !supportedRules.Has(string(groupOptions.Rule)) {
allErrs = append(allErrs, field.NotSupported(fldPath.Child("rule"), groupOptions.Rule, supportedRules.List()))
}
for idx, rng := range groupOptions.Ranges {
allErrs = append(allErrs, validateGroupIDRange(fldPath.Child("ranges").Index(idx), rng)...)
}
return allErrs
}
// validatePodSecurityPolicyVolumes validates the volume fields of PodSecurityPolicy.
func validatePodSecurityPolicyVolumes(fldPath *field.Path, volumes []policy.FSType) field.ErrorList {
allErrs := field.ErrorList{}
allowed := psputil.GetAllFSTypesAsSet()
// add in the * value since that is a pseudo type that is not included by default
allowed.Insert(string(policy.All))
for _, v := range volumes {
if !allowed.Has(string(v)) {
allErrs = append(allErrs, field.NotSupported(fldPath.Child("volumes"), v, allowed.List()))
}
}
return allErrs
}
// validatePSPDefaultAllowPrivilegeEscalation validates the DefaultAllowPrivilegeEscalation field against the AllowPrivilegeEscalation field of a PodSecurityPolicy.
func validatePSPDefaultAllowPrivilegeEscalation(fldPath *field.Path, defaultAllowPrivilegeEscalation *bool, allowPrivilegeEscalation bool) field.ErrorList {
allErrs := field.ErrorList{}
if defaultAllowPrivilegeEscalation != nil && *defaultAllowPrivilegeEscalation && !allowPrivilegeEscalation {
allErrs = append(allErrs, field.Invalid(fldPath, defaultAllowPrivilegeEscalation, "Cannot set DefaultAllowPrivilegeEscalation to true without also setting AllowPrivilegeEscalation to true"))
}
return allErrs
}
// validatePSPAllowedProcMountTypes validates the DefaultAllowPrivilegeEscalation field against the AllowPrivilegeEscalation field of a PodSecurityPolicy.
func validatePSPAllowedProcMountTypes(fldPath *field.Path, allowedProcMountTypes []core.ProcMountType) field.ErrorList {
allErrs := field.ErrorList{}
for i, procMountType := range allowedProcMountTypes {
if err := apivalidation.ValidateProcMountType(fldPath.Index(i), procMountType); err != nil {
allErrs = append(allErrs, err)
}
}
return allErrs
}
const sysctlPatternSegmentFmt string = "([a-z0-9][-_a-z0-9]*)?[a-z0-9*]"
// SysctlPatternFmt is a regex used for matching valid sysctl patterns.
const SysctlPatternFmt string = "(" + apivalidation.SysctlSegmentFmt + "\\.)*" + sysctlPatternSegmentFmt
var sysctlPatternRegexp = regexp.MustCompile("^" + SysctlPatternFmt + "$")
// IsValidSysctlPattern checks if name is a valid sysctl pattern.
func IsValidSysctlPattern(name string) bool {
if len(name) > apivalidation.SysctlMaxLength {
return false
}
return sysctlPatternRegexp.MatchString(name)
}
func validatePodSecurityPolicySysctlListsDoNotOverlap(allowedSysctlsFldPath, forbiddenSysctlsFldPath *field.Path, allowedUnsafeSysctls, forbiddenSysctls []string) field.ErrorList {
allErrs := field.ErrorList{}
for i, allowedSysctl := range allowedUnsafeSysctls {
isAllowedSysctlPattern := false
allowedSysctlPrefix := ""
if strings.HasSuffix(allowedSysctl, "*") {
isAllowedSysctlPattern = true
allowedSysctlPrefix = strings.TrimSuffix(allowedSysctl, "*")
}
for j, forbiddenSysctl := range forbiddenSysctls {
isForbiddenSysctlPattern := false
forbiddenSysctlPrefix := ""
if strings.HasSuffix(forbiddenSysctl, "*") {
isForbiddenSysctlPattern = true
forbiddenSysctlPrefix = strings.TrimSuffix(forbiddenSysctl, "*")
}
switch {
case isAllowedSysctlPattern && isForbiddenSysctlPattern:
if strings.HasPrefix(allowedSysctlPrefix, forbiddenSysctlPrefix) {
allErrs = append(allErrs, field.Invalid(allowedSysctlsFldPath.Index(i), allowedUnsafeSysctls[i], fmt.Sprintf("sysctl overlaps with %v", forbiddenSysctl)))
} else if strings.HasPrefix(forbiddenSysctlPrefix, allowedSysctlPrefix) {
allErrs = append(allErrs, field.Invalid(forbiddenSysctlsFldPath.Index(j), forbiddenSysctls[j], fmt.Sprintf("sysctl overlaps with %v", allowedSysctl)))
}
case isAllowedSysctlPattern:
if strings.HasPrefix(forbiddenSysctl, allowedSysctlPrefix) {
allErrs = append(allErrs, field.Invalid(forbiddenSysctlsFldPath.Index(j), forbiddenSysctls[j], fmt.Sprintf("sysctl overlaps with %v", allowedSysctl)))
}
case isForbiddenSysctlPattern:
if strings.HasPrefix(allowedSysctl, forbiddenSysctlPrefix) {
allErrs = append(allErrs, field.Invalid(allowedSysctlsFldPath.Index(i), allowedUnsafeSysctls[i], fmt.Sprintf("sysctl overlaps with %v", forbiddenSysctl)))
}
default:
if allowedSysctl == forbiddenSysctl {
allErrs = append(allErrs, field.Invalid(allowedSysctlsFldPath.Index(i), allowedUnsafeSysctls[i], fmt.Sprintf("sysctl overlaps with %v", forbiddenSysctl)))
}
}
}
}
return allErrs
}
// validatePodSecurityPolicySysctls validates the sysctls fields of PodSecurityPolicy.
func validatePodSecurityPolicySysctls(fldPath *field.Path, sysctls []string) field.ErrorList {
allErrs := field.ErrorList{}
if len(sysctls) == 0 {
return allErrs
}
coversAll := false
for i, s := range sysctls {
if len(s) == 0 {
allErrs = append(allErrs, field.Invalid(fldPath.Index(i), sysctls[i], fmt.Sprintf("empty sysctl not allowed")))
} else if !IsValidSysctlPattern(string(s)) {
allErrs = append(
allErrs,
field.Invalid(fldPath.Index(i), sysctls[i], fmt.Sprintf("must have at most %d characters and match regex %s",
apivalidation.SysctlMaxLength,
SysctlPatternFmt,
)),
)
} else if s[0] == '*' {
coversAll = true
}
}
if coversAll && len(sysctls) > 1 {
allErrs = append(allErrs, field.Forbidden(fldPath.Child("items"), fmt.Sprintf("if '*' is present, must not specify other sysctls")))
}
return allErrs
}
func validateUserIDRange(fldPath *field.Path, rng policy.IDRange) field.ErrorList {
return validateIDRanges(fldPath, rng.Min, rng.Max)
}
func validateGroupIDRange(fldPath *field.Path, rng policy.IDRange) field.ErrorList {
return validateIDRanges(fldPath, rng.Min, rng.Max)
}
// validateIDRanges ensures the range is valid.
func validateIDRanges(fldPath *field.Path, min, max int64) field.ErrorList {
allErrs := field.ErrorList{}
// if 0 <= Min <= Max then we do not need to validate max. It is always greater than or
// equal to 0 and Min.
if min < 0 {
allErrs = append(allErrs, field.Invalid(fldPath.Child("min"), min, "min cannot be negative"))
}
if max < 0 {
allErrs = append(allErrs, field.Invalid(fldPath.Child("max"), max, "max cannot be negative"))
}
if min > max {
allErrs = append(allErrs, field.Invalid(fldPath.Child("min"), min, "min cannot be greater than max"))
}
return allErrs
}
// validatePSPCapsAgainstDrops ensures an allowed cap is not listed in the required drops.
func validatePSPCapsAgainstDrops(requiredDrops []core.Capability, capsToCheck []core.Capability, fldPath *field.Path) field.ErrorList {
allErrs := field.ErrorList{}
if requiredDrops == nil {
return allErrs
}
for _, cap := range capsToCheck {
if hasCap(cap, requiredDrops) {
allErrs = append(allErrs, field.Invalid(fldPath, cap,
fmt.Sprintf("capability is listed in %s and requiredDropCapabilities", fldPath.String())))
}
}
return allErrs
}
// validateRuntimeClassStrategy ensures all the RuntimeClass restrictions are valid.
func validateRuntimeClassStrategy(fldPath *field.Path, rc *policy.RuntimeClassStrategyOptions) field.ErrorList {
if rc == nil {
return nil
}
var allErrs field.ErrorList
allowed := map[string]bool{}
for i, name := range rc.AllowedRuntimeClassNames {
if name != policy.AllowAllRuntimeClassNames {
allErrs = append(allErrs, apivalidation.ValidateRuntimeClassName(name, fldPath.Child("allowedRuntimeClassNames").Index(i))...)
}
if allowed[name] {
allErrs = append(allErrs, field.Duplicate(fldPath.Child("allowedRuntimeClassNames").Index(i), name))
}
allowed[name] = true
}
if rc.DefaultRuntimeClassName != nil {
allErrs = append(allErrs, apivalidation.ValidateRuntimeClassName(*rc.DefaultRuntimeClassName, fldPath.Child("defaultRuntimeClassName"))...)
if !allowed[*rc.DefaultRuntimeClassName] && !allowed[policy.AllowAllRuntimeClassNames] {
allErrs = append(allErrs, field.Required(fldPath.Child("allowedRuntimeClassNames"),
fmt.Sprintf("default %q must be allowed", *rc.DefaultRuntimeClassName)))
}
}
if allowed[policy.AllowAllRuntimeClassNames] && len(rc.AllowedRuntimeClassNames) > 1 {
allErrs = append(allErrs, field.Invalid(fldPath.Child("allowedRuntimeClassNames"), rc.AllowedRuntimeClassNames, "if '*' is present, must not specify other RuntimeClass names"))
}
return allErrs
}
// ValidatePodSecurityPolicyUpdate validates a PSP for updates.
func ValidatePodSecurityPolicyUpdate(old *policy.PodSecurityPolicy, new *policy.PodSecurityPolicy) field.ErrorList {
allErrs := field.ErrorList{}
allErrs = append(allErrs, apivalidation.ValidateObjectMetaUpdate(&new.ObjectMeta, &old.ObjectMeta, field.NewPath("metadata"))...)
allErrs = append(allErrs, ValidatePodSecurityPolicySpecificAnnotations(new.Annotations, field.NewPath("metadata").Child("annotations"))...)
allErrs = append(allErrs, ValidatePodSecurityPolicySpec(&new.Spec, field.NewPath("spec"))...)
return allErrs
}
// hasCap checks for needle in haystack.
func hasCap(needle core.Capability, haystack []core.Capability) bool {
for _, c := range haystack {
if needle == c {
return true
}
}
return false
}

View File

@ -69,7 +69,7 @@ type PriorityClass struct {
Description string
// PreemptionPolicy it the Policy for preempting pods with lower priority.
// This field is alpha-level and is only honored by servers that enable the NonPreemptingPriority feature.
// This field is beta-level.
// +optional
PreemptionPolicy *core.PreemptionPolicy
}

View File

@ -1,26 +0,0 @@
package(default_visibility = ["//visibility:public"])
load(
"@io_bazel_rules_go//go:def.bzl",
"go_library",
)
go_library(
name = "go_default_library",
srcs = ["helpers.go"],
importpath = "k8s.io/kubernetes/pkg/apis/storage/v1/util",
deps = ["//staging/src/k8s.io/apimachinery/pkg/apis/meta/v1:go_default_library"],
)
filegroup(
name = "package-srcs",
srcs = glob(["**"]),
tags = ["automanaged"],
visibility = ["//visibility:private"],
)
filegroup(
name = "all-srcs",
srcs = [":package-srcs"],
tags = ["automanaged"],
)

View File

@ -1,57 +0,0 @@
/*
Copyright 2016 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 util
import (
metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
)
// IsDefaultStorageClassAnnotation represents a StorageClass annotation that
// marks a class as the default StorageClass
const IsDefaultStorageClassAnnotation = "storageclass.kubernetes.io/is-default-class"
// BetaIsDefaultStorageClassAnnotation is the beta version of IsDefaultStorageClassAnnotation.
// TODO: remove Beta when no longer used
const BetaIsDefaultStorageClassAnnotation = "storageclass.beta.kubernetes.io/is-default-class"
// IsDefaultAnnotationText returns a pretty Yes/No String if
// the annotation is set
// TODO: remove Beta when no longer needed
func IsDefaultAnnotationText(obj metav1.ObjectMeta) string {
if obj.Annotations[IsDefaultStorageClassAnnotation] == "true" {
return "Yes"
}
if obj.Annotations[BetaIsDefaultStorageClassAnnotation] == "true" {
return "Yes"
}
return "No"
}
// IsDefaultAnnotation returns a boolean if
// the annotation is set
// TODO: remove Beta when no longer needed
func IsDefaultAnnotation(obj metav1.ObjectMeta) bool {
if obj.Annotations[IsDefaultStorageClassAnnotation] == "true" {
return true
}
if obj.Annotations[BetaIsDefaultStorageClassAnnotation] == "true" {
return true
}
return false
}

View File

@ -11,7 +11,10 @@ go_library(
"doc.go",
"ports.go",
],
importpath = "k8s.io/kubernetes/pkg/master/ports",
importpath = "k8s.io/kubernetes/pkg/cluster/ports",
deps = [
"//staging/src/k8s.io/cloud-provider:go_default_library",
],
)
filegroup(

View File

@ -16,4 +16,4 @@ limitations under the License.
// Package ports defines ports used by various pieces of the kubernetes
// infrastructure.
package ports // import "k8s.io/kubernetes/pkg/master/ports"
package ports // import "k8s.io/kubernetes/pkg/cluster/ports"

View File

@ -16,6 +16,10 @@ limitations under the License.
package ports
import (
"k8s.io/cloud-provider"
)
const (
// ProxyStatusPort is the default port for the proxy metrics server.
// May be overridden by a flag at startup.
@ -23,18 +27,10 @@ const (
// KubeletPort is the default port for the kubelet server on each host machine.
// May be overridden by a flag at startup.
KubeletPort = 10250
// InsecureSchedulerPort is the default port for the scheduler status server.
// May be overridden by a flag at startup.
// Deprecated: use the secure KubeSchedulerPort instead.
InsecureSchedulerPort = 10251
// InsecureKubeControllerManagerPort is the default port for the controller manager status server.
// May be overridden by a flag at startup.
// Deprecated: use the secure KubeControllerManagerPort instead.
InsecureKubeControllerManagerPort = 10252
// InsecureCloudControllerManagerPort is the default port for the cloud controller manager server.
// This value may be overridden by a flag at startup.
// Deprecated: use the secure CloudControllerManagerPort instead.
InsecureCloudControllerManagerPort = 10253
// KubeletReadOnlyPort exposes basic read-only services from the kubelet.
// May be overridden by a flag at startup.
// This is necessary for heapster to collect monitoring stats from the kubelet
@ -49,9 +45,5 @@ const (
KubeControllerManagerPort = 10257
// CloudControllerManagerPort is the default port for the cloud controller manager server.
// This value may be overridden by a flag at startup.
CloudControllerManagerPort = 10258
// KubeSchedulerPort is the default port for the scheduler status server.
// May be overridden by a flag at startup.
KubeSchedulerPort = 10259
CloudControllerManagerPort = cloudprovider.CloudControllerManagerPort
)

View File

@ -1,16 +0,0 @@
{
"Rules": [
{
"SelectorRegexp": "k8s[.]io/kubernetes/pkg/client/unversioned$",
"ForbiddenPrefixes": [
"k8s.io/kubernetes/pkg/client/unversioned"
]
},
{
"SelectorRegexp": "k8s[.]io/kubernetes/pkg/client/unversioned/testclient$",
"ForbiddenPrefixes": [
"k8s.io/kubernetes/pkg/client/unversioned/testclient"
]
}
]
}

View File

@ -1,9 +1,53 @@
package(default_visibility = ["//visibility:public"])
load("@io_bazel_rules_go//go:def.bzl", "go_library", "go_test")
load(
"@io_bazel_rules_go//go:def.bzl",
"go_library",
"go_test",
go_library(
name = "go_default_library",
srcs = [
"client_builder_dynamic.go",
"controller_ref_manager.go",
"controller_utils.go",
"doc.go",
"lookup_cache.go",
],
importpath = "k8s.io/kubernetes/pkg/controller",
visibility = ["//visibility:public"],
deps = [
"//pkg/api/v1/pod:go_default_library",
"//pkg/apis/core/install:go_default_library",
"//pkg/apis/core/validation:go_default_library",
"//pkg/util/hash:go_default_library",
"//pkg/util/taints:go_default_library",
"//staging/src/k8s.io/api/apps/v1:go_default_library",
"//staging/src/k8s.io/api/authentication/v1:go_default_library",
"//staging/src/k8s.io/api/core/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/api/errors:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/api/meta:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/apis/meta/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/labels:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/runtime:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/runtime/schema:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/types:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/clock:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/errors:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/rand:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/sets:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/strategicpatch:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/wait:go_default_library",
"//staging/src/k8s.io/apiserver/pkg/authentication/serviceaccount:go_default_library",
"//staging/src/k8s.io/client-go/kubernetes:go_default_library",
"//staging/src/k8s.io/client-go/kubernetes/typed/core/v1:go_default_library",
"//staging/src/k8s.io/client-go/rest:go_default_library",
"//staging/src/k8s.io/client-go/tools/cache:go_default_library",
"//staging/src/k8s.io/client-go/tools/record:go_default_library",
"//staging/src/k8s.io/client-go/transport:go_default_library",
"//staging/src/k8s.io/client-go/util/retry:go_default_library",
"//staging/src/k8s.io/controller-manager/pkg/clientbuilder:go_default_library",
"//vendor/github.com/golang/groupcache/lru:go_default_library",
"//vendor/golang.org/x/oauth2:go_default_library",
"//vendor/k8s.io/klog/v2:go_default_library",
"//vendor/k8s.io/utils/integer:go_default_library",
"//vendor/k8s.io/utils/pointer:go_default_library",
],
)
go_test(
@ -20,6 +64,7 @@ go_test(
"//staging/src/k8s.io/api/apps/v1:go_default_library",
"//staging/src/k8s.io/api/core/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/api/equality:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/api/errors:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/apis/meta/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/labels:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/runtime:go_default_library",
@ -39,65 +84,6 @@ go_test(
],
)
go_library(
name = "go_default_library",
srcs = [
"client_builder.go",
"client_builder_dynamic.go",
"controller_ref_manager.go",
"controller_utils.go",
"doc.go",
"informer_factory.go",
"lookup_cache.go",
],
importpath = "k8s.io/kubernetes/pkg/controller",
deps = [
"//pkg/api/legacyscheme:go_default_library",
"//pkg/api/v1/pod:go_default_library",
"//pkg/apis/core:go_default_library",
"//pkg/apis/core/install:go_default_library",
"//pkg/apis/core/validation:go_default_library",
"//pkg/serviceaccount:go_default_library",
"//pkg/util/hash:go_default_library",
"//pkg/util/taints:go_default_library",
"//staging/src/k8s.io/api/apps/v1:go_default_library",
"//staging/src/k8s.io/api/authentication/v1:go_default_library",
"//staging/src/k8s.io/api/core/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/api/errors:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/api/meta:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/apis/meta/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/fields:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/labels:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/runtime:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/runtime/schema:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/types:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/clock:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/errors:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/rand:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/sets:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/strategicpatch:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/wait:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/watch:go_default_library",
"//staging/src/k8s.io/apiserver/pkg/authentication/serviceaccount:go_default_library",
"//staging/src/k8s.io/client-go/informers:go_default_library",
"//staging/src/k8s.io/client-go/kubernetes:go_default_library",
"//staging/src/k8s.io/client-go/kubernetes/typed/authentication/v1:go_default_library",
"//staging/src/k8s.io/client-go/kubernetes/typed/core/v1:go_default_library",
"//staging/src/k8s.io/client-go/metadata/metadatainformer:go_default_library",
"//staging/src/k8s.io/client-go/rest:go_default_library",
"//staging/src/k8s.io/client-go/tools/cache:go_default_library",
"//staging/src/k8s.io/client-go/tools/record:go_default_library",
"//staging/src/k8s.io/client-go/tools/watch:go_default_library",
"//staging/src/k8s.io/client-go/transport:go_default_library",
"//staging/src/k8s.io/client-go/util/retry:go_default_library",
"//vendor/github.com/golang/groupcache/lru:go_default_library",
"//vendor/golang.org/x/oauth2:go_default_library",
"//vendor/k8s.io/klog:go_default_library",
"//vendor/k8s.io/utils/integer:go_default_library",
"//vendor/k8s.io/utils/pointer:go_default_library",
],
)
filegroup(
name = "package-srcs",
srcs = glob(["**"]),
@ -112,7 +98,6 @@ filegroup(
"//pkg/controller/apis/config:all-srcs",
"//pkg/controller/bootstrap:all-srcs",
"//pkg/controller/certificates:all-srcs",
"//pkg/controller/cloud:all-srcs",
"//pkg/controller/clusterroleaggregation:all-srcs",
"//pkg/controller/cronjob:all-srcs",
"//pkg/controller/daemon:all-srcs",
@ -120,6 +105,7 @@ filegroup(
"//pkg/controller/disruption:all-srcs",
"//pkg/controller/endpoint:all-srcs",
"//pkg/controller/endpointslice:all-srcs",
"//pkg/controller/endpointslicemirroring:all-srcs",
"//pkg/controller/garbagecollector:all-srcs",
"//pkg/controller/history:all-srcs",
"//pkg/controller/job:all-srcs",
@ -131,16 +117,17 @@ filegroup(
"//pkg/controller/replicaset:all-srcs",
"//pkg/controller/replication:all-srcs",
"//pkg/controller/resourcequota:all-srcs",
"//pkg/controller/route:all-srcs",
"//pkg/controller/service:all-srcs",
"//pkg/controller/serviceaccount:all-srcs",
"//pkg/controller/statefulset:all-srcs",
"//pkg/controller/storageversiongc:all-srcs",
"//pkg/controller/testutil:all-srcs",
"//pkg/controller/ttl:all-srcs",
"//pkg/controller/ttlafterfinished:all-srcs",
"//pkg/controller/util/endpoint:all-srcs",
"//pkg/controller/util/node:all-srcs",
"//pkg/controller/volume/attachdetach:all-srcs",
"//pkg/controller/volume/common:all-srcs",
"//pkg/controller/volume/ephemeral:all-srcs",
"//pkg/controller/volume/events:all-srcs",
"//pkg/controller/volume/expand:all-srcs",
"//pkg/controller/volume/persistentvolume:all-srcs",
@ -150,4 +137,5 @@ filegroup(
"//pkg/controller/volume/scheduling:all-srcs",
],
tags = ["automanaged"],
visibility = ["//visibility:public"],
)

View File

@ -5,8 +5,11 @@ approvers:
- derekwaynecarr
- mikedanese
- janetkuo
- cheftako
- sig-apps-approvers
reviewers:
- deads2k
- cheftako
- sig-apps-reviewers
labels:
- sig/apps

View File

@ -1,241 +0,0 @@
/*
Copyright 2016 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 controller
import (
"context"
"fmt"
"time"
v1authenticationapi "k8s.io/api/authentication/v1"
v1 "k8s.io/api/core/v1"
apierrors "k8s.io/apimachinery/pkg/api/errors"
metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
"k8s.io/apimachinery/pkg/fields"
"k8s.io/apimachinery/pkg/runtime"
"k8s.io/apimachinery/pkg/watch"
apiserverserviceaccount "k8s.io/apiserver/pkg/authentication/serviceaccount"
clientset "k8s.io/client-go/kubernetes"
v1authentication "k8s.io/client-go/kubernetes/typed/authentication/v1"
v1core "k8s.io/client-go/kubernetes/typed/core/v1"
restclient "k8s.io/client-go/rest"
"k8s.io/client-go/tools/cache"
watchtools "k8s.io/client-go/tools/watch"
"k8s.io/klog"
"k8s.io/kubernetes/pkg/api/legacyscheme"
api "k8s.io/kubernetes/pkg/apis/core"
"k8s.io/kubernetes/pkg/serviceaccount"
)
// ControllerClientBuilder allows you to get clients and configs for controllers
// Please note a copy also exists in staging/src/k8s.io/cloud-provider/cloud.go
// TODO: Extract this into a separate controller utilities repo (issues/68947)
type ControllerClientBuilder interface {
Config(name string) (*restclient.Config, error)
ConfigOrDie(name string) *restclient.Config
Client(name string) (clientset.Interface, error)
ClientOrDie(name string) clientset.Interface
}
// SimpleControllerClientBuilder returns a fixed client with different user agents
type SimpleControllerClientBuilder struct {
// ClientConfig is a skeleton config to clone and use as the basis for each controller client
ClientConfig *restclient.Config
}
func (b SimpleControllerClientBuilder) Config(name string) (*restclient.Config, error) {
clientConfig := *b.ClientConfig
return restclient.AddUserAgent(&clientConfig, name), nil
}
func (b SimpleControllerClientBuilder) ConfigOrDie(name string) *restclient.Config {
clientConfig, err := b.Config(name)
if err != nil {
klog.Fatal(err)
}
return clientConfig
}
func (b SimpleControllerClientBuilder) Client(name string) (clientset.Interface, error) {
clientConfig, err := b.Config(name)
if err != nil {
return nil, err
}
return clientset.NewForConfig(clientConfig)
}
func (b SimpleControllerClientBuilder) ClientOrDie(name string) clientset.Interface {
client, err := b.Client(name)
if err != nil {
klog.Fatal(err)
}
return client
}
// SAControllerClientBuilder is a ControllerClientBuilder that returns clients identifying as
// service accounts
type SAControllerClientBuilder struct {
// ClientConfig is a skeleton config to clone and use as the basis for each controller client
ClientConfig *restclient.Config
// CoreClient is used to provision service accounts if needed and watch for their associated tokens
// to construct a controller client
CoreClient v1core.CoreV1Interface
// AuthenticationClient is used to check API tokens to make sure they are valid before
// building a controller client from them
AuthenticationClient v1authentication.AuthenticationV1Interface
// Namespace is the namespace used to host the service accounts that will back the
// controllers. It must be highly privileged namespace which normal users cannot inspect.
Namespace string
}
// config returns a complete clientConfig for constructing clients. This is separate in anticipation of composition
// which means that not all clientsets are known here
func (b SAControllerClientBuilder) Config(name string) (*restclient.Config, error) {
sa, err := getOrCreateServiceAccount(b.CoreClient, b.Namespace, name)
if err != nil {
return nil, err
}
var clientConfig *restclient.Config
fieldSelector := fields.SelectorFromSet(map[string]string{
api.SecretTypeField: string(v1.SecretTypeServiceAccountToken),
}).String()
lw := &cache.ListWatch{
ListFunc: func(options metav1.ListOptions) (runtime.Object, error) {
options.FieldSelector = fieldSelector
return b.CoreClient.Secrets(b.Namespace).List(context.TODO(), options)
},
WatchFunc: func(options metav1.ListOptions) (watch.Interface, error) {
options.FieldSelector = fieldSelector
return b.CoreClient.Secrets(b.Namespace).Watch(context.TODO(), options)
},
}
ctx, cancel := context.WithTimeout(context.Background(), 30*time.Second)
defer cancel()
_, err = watchtools.UntilWithSync(ctx, lw, &v1.Secret{}, nil,
func(event watch.Event) (bool, error) {
switch event.Type {
case watch.Deleted:
return false, nil
case watch.Error:
return false, fmt.Errorf("error watching")
case watch.Added, watch.Modified:
secret, ok := event.Object.(*v1.Secret)
if !ok {
return false, fmt.Errorf("unexpected object type: %T", event.Object)
}
if !serviceaccount.IsServiceAccountToken(secret, sa) {
return false, nil
}
if len(secret.Data[v1.ServiceAccountTokenKey]) == 0 {
return false, nil
}
validConfig, valid, err := b.getAuthenticatedConfig(sa, string(secret.Data[v1.ServiceAccountTokenKey]))
if err != nil {
klog.Warningf("error validating API token for %s/%s in secret %s: %v", sa.Namespace, sa.Name, secret.Name, err)
// continue watching for good tokens
return false, nil
}
if !valid {
klog.Warningf("secret %s contained an invalid API token for %s/%s", secret.Name, sa.Namespace, sa.Name)
// try to delete the secret containing the invalid token
if err := b.CoreClient.Secrets(secret.Namespace).Delete(context.TODO(), secret.Name, metav1.DeleteOptions{}); err != nil && !apierrors.IsNotFound(err) {
klog.Warningf("error deleting secret %s containing invalid API token for %s/%s: %v", secret.Name, sa.Namespace, sa.Name, err)
}
// continue watching for good tokens
return false, nil
}
clientConfig = validConfig
return true, nil
default:
return false, fmt.Errorf("unexpected event type: %v", event.Type)
}
})
if err != nil {
return nil, fmt.Errorf("unable to get token for service account: %v", err)
}
return clientConfig, nil
}
func (b SAControllerClientBuilder) getAuthenticatedConfig(sa *v1.ServiceAccount, token string) (*restclient.Config, bool, error) {
username := apiserverserviceaccount.MakeUsername(sa.Namespace, sa.Name)
clientConfig := restclient.AnonymousClientConfig(b.ClientConfig)
clientConfig.BearerToken = token
restclient.AddUserAgent(clientConfig, username)
// Try token review first
tokenReview := &v1authenticationapi.TokenReview{Spec: v1authenticationapi.TokenReviewSpec{Token: token}}
if tokenResult, err := b.AuthenticationClient.TokenReviews().Create(context.TODO(), tokenReview, metav1.CreateOptions{}); err == nil {
if !tokenResult.Status.Authenticated {
klog.Warningf("Token for %s/%s did not authenticate correctly", sa.Namespace, sa.Name)
return nil, false, nil
}
if tokenResult.Status.User.Username != username {
klog.Warningf("Token for %s/%s authenticated as unexpected username: %s", sa.Namespace, sa.Name, tokenResult.Status.User.Username)
return nil, false, nil
}
klog.V(4).Infof("Verified credential for %s/%s", sa.Namespace, sa.Name)
return clientConfig, true, nil
}
// If we couldn't run the token review, the API might be disabled or we might not have permission.
// Try to make a request to /apis with the token. If we get a 401 we should consider the token invalid.
clientConfigCopy := *clientConfig
clientConfigCopy.NegotiatedSerializer = legacyscheme.Codecs
client, err := restclient.UnversionedRESTClientFor(&clientConfigCopy)
if err != nil {
return nil, false, err
}
err = client.Get().AbsPath("/apis").Do(context.TODO()).Error()
if apierrors.IsUnauthorized(err) {
klog.Warningf("Token for %s/%s did not authenticate correctly: %v", sa.Namespace, sa.Name, err)
return nil, false, nil
}
return clientConfig, true, nil
}
func (b SAControllerClientBuilder) ConfigOrDie(name string) *restclient.Config {
clientConfig, err := b.Config(name)
if err != nil {
klog.Fatal(err)
}
return clientConfig
}
func (b SAControllerClientBuilder) Client(name string) (clientset.Interface, error) {
clientConfig, err := b.Config(name)
if err != nil {
return nil, err
}
return clientset.NewForConfig(clientConfig)
}
func (b SAControllerClientBuilder) ClientOrDie(name string) clientset.Interface {
client, err := b.Client(name)
if err != nil {
klog.Fatal(err)
}
return client
}

View File

@ -24,7 +24,6 @@ import (
"time"
"golang.org/x/oauth2"
v1authenticationapi "k8s.io/api/authentication/v1"
metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
"k8s.io/apimachinery/pkg/util/clock"
@ -34,7 +33,8 @@ import (
v1core "k8s.io/client-go/kubernetes/typed/core/v1"
restclient "k8s.io/client-go/rest"
"k8s.io/client-go/transport"
"k8s.io/klog"
"k8s.io/controller-manager/pkg/clientbuilder"
"k8s.io/klog/v2"
utilpointer "k8s.io/utils/pointer"
)
@ -73,7 +73,7 @@ type DynamicControllerClientBuilder struct {
clock clock.Clock
}
func NewDynamicClientBuilder(clientConfig *restclient.Config, coreClient v1core.CoreV1Interface, ns string) ControllerClientBuilder {
func NewDynamicClientBuilder(clientConfig *restclient.Config, coreClient v1core.CoreV1Interface, ns string) clientbuilder.ControllerClientBuilder {
builder := &DynamicControllerClientBuilder{
ClientConfig: clientConfig,
CoreClient: coreClient,
@ -87,7 +87,7 @@ func NewDynamicClientBuilder(clientConfig *restclient.Config, coreClient v1core.
}
// this function only for test purpose, don't call it
func NewTestDynamicClientBuilder(clientConfig *restclient.Config, coreClient v1core.CoreV1Interface, ns string, expirationSeconds int64, leewayPercent int) ControllerClientBuilder {
func NewTestDynamicClientBuilder(clientConfig *restclient.Config, coreClient v1core.CoreV1Interface, ns string, expirationSeconds int64, leewayPercent int) clientbuilder.ControllerClientBuilder {
builder := &DynamicControllerClientBuilder{
ClientConfig: clientConfig,
CoreClient: coreClient,

View File

@ -29,7 +29,7 @@ import (
"k8s.io/apimachinery/pkg/runtime/schema"
"k8s.io/apimachinery/pkg/types"
utilerrors "k8s.io/apimachinery/pkg/util/errors"
"k8s.io/klog"
"k8s.io/klog/v2"
)
type BaseControllerRefManager struct {

View File

@ -51,7 +51,7 @@ import (
taintutils "k8s.io/kubernetes/pkg/util/taints"
"k8s.io/utils/integer"
"k8s.io/klog"
"k8s.io/klog/v2"
)
const (
@ -601,8 +601,12 @@ func (r RealPodControl) DeletePod(namespace string, podID string, object runtime
if err != nil {
return fmt.Errorf("object does not have ObjectMeta, %v", err)
}
klog.V(2).Infof("Controller %v deleting pod %v/%v", accessor.GetName(), namespace, podID)
if err := r.KubeClient.CoreV1().Pods(namespace).Delete(context.TODO(), podID, metav1.DeleteOptions{}); err != nil && !apierrors.IsNotFound(err) {
klog.V(2).InfoS("Deleting pod", "controller", accessor.GetName(), "pod", klog.KRef(namespace, podID))
if err := r.KubeClient.CoreV1().Pods(namespace).Delete(context.TODO(), podID, metav1.DeleteOptions{}); err != nil {
if apierrors.IsNotFound(err) {
klog.V(4).Infof("pod %v/%v has already been deleted.", namespace, podID)
return err
}
r.Recorder.Eventf(object, v1.EventTypeWarning, FailedDeletePodReason, "Error deleting: %v", err)
return fmt.Errorf("unable to delete pods: %v", err)
}

View File

@ -25,7 +25,7 @@ go_library(
"//staging/src/k8s.io/apimachinery/pkg/util/wait:go_default_library",
"//staging/src/k8s.io/client-go/kubernetes/typed/apps/v1:go_default_library",
"//staging/src/k8s.io/client-go/listers/apps/v1:go_default_library",
"//vendor/k8s.io/klog:go_default_library",
"//vendor/k8s.io/klog/v2:go_default_library",
"//vendor/k8s.io/utils/integer:go_default_library",
],
)

View File

@ -37,7 +37,7 @@ import (
"k8s.io/apimachinery/pkg/util/wait"
appsclient "k8s.io/client-go/kubernetes/typed/apps/v1"
appslisters "k8s.io/client-go/listers/apps/v1"
"k8s.io/klog"
"k8s.io/klog/v2"
"k8s.io/kubernetes/pkg/controller"
labelsutil "k8s.io/kubernetes/pkg/util/labels"
"k8s.io/utils/integer"
@ -817,7 +817,7 @@ func NewRSNewReplicas(deployment *apps.Deployment, allRSs []*apps.ReplicaSet, ne
switch deployment.Spec.Strategy.Type {
case apps.RollingUpdateDeploymentStrategyType:
// Check if we can scale up.
maxSurge, err := intstrutil.GetValueFromIntOrPercent(deployment.Spec.Strategy.RollingUpdate.MaxSurge, int(*(deployment.Spec.Replicas)), true)
maxSurge, err := intstrutil.GetScaledValueFromIntOrPercent(deployment.Spec.Strategy.RollingUpdate.MaxSurge, int(*(deployment.Spec.Replicas)), true)
if err != nil {
return 0, err
}
@ -881,11 +881,11 @@ func WaitForObservedDeployment(getDeploymentFunc func() (*apps.Deployment, error
// 2 desired, max unavailable 0%, surge 1% - should scale new(+1), then old(-1), then new(+1), then old(-1)
// 1 desired, max unavailable 0%, surge 1% - should scale new(+1), then old(-1)
func ResolveFenceposts(maxSurge, maxUnavailable *intstrutil.IntOrString, desired int32) (int32, int32, error) {
surge, err := intstrutil.GetValueFromIntOrPercent(intstrutil.ValueOrDefault(maxSurge, intstrutil.FromInt(0)), int(desired), true)
surge, err := intstrutil.GetScaledValueFromIntOrPercent(intstrutil.ValueOrDefault(maxSurge, intstrutil.FromInt(0)), int(desired), true)
if err != nil {
return 0, 0, err
}
unavailable, err := intstrutil.GetValueFromIntOrPercent(intstrutil.ValueOrDefault(maxUnavailable, intstrutil.FromInt(0)), int(desired), false)
unavailable, err := intstrutil.GetScaledValueFromIntOrPercent(intstrutil.ValueOrDefault(maxUnavailable, intstrutil.FromInt(0)), int(desired), false)
if err != nil {
return 0, 0, err
}

View File

@ -1,56 +0,0 @@
/*
Copyright 2019 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 controller
import (
"k8s.io/apimachinery/pkg/runtime/schema"
"k8s.io/client-go/informers"
"k8s.io/client-go/metadata/metadatainformer"
)
// InformerFactory creates informers for each group version resource.
type InformerFactory interface {
ForResource(resource schema.GroupVersionResource) (informers.GenericInformer, error)
Start(stopCh <-chan struct{})
}
type informerFactory struct {
typedInformerFactory informers.SharedInformerFactory
metadataInformerFactory metadatainformer.SharedInformerFactory
}
func (i *informerFactory) ForResource(resource schema.GroupVersionResource) (informers.GenericInformer, error) {
informer, err := i.typedInformerFactory.ForResource(resource)
if err != nil {
return i.metadataInformerFactory.ForResource(resource), nil
}
return informer, nil
}
func (i *informerFactory) Start(stopCh <-chan struct{}) {
i.typedInformerFactory.Start(stopCh)
i.metadataInformerFactory.Start(stopCh)
}
// NewInformerFactory creates a new InformerFactory which works with both typed
// resources and metadata-only resources
func NewInformerFactory(typedInformerFactory informers.SharedInformerFactory, metadataInformerFactory metadatainformer.SharedInformerFactory) InformerFactory {
return &informerFactory{
typedInformerFactory: typedInformerFactory,
metadataInformerFactory: metadataInformerFactory,
}
}

View File

@ -1,100 +0,0 @@
load("@io_bazel_rules_go//go:def.bzl", "go_library", "go_test")
go_library(
name = "go_default_library",
srcs = [
"metrics.go",
"node_lifecycle_controller.go",
],
importpath = "k8s.io/kubernetes/pkg/controller/nodelifecycle",
visibility = ["//visibility:public"],
deps = [
"//pkg/controller:go_default_library",
"//pkg/controller/nodelifecycle/scheduler:go_default_library",
"//pkg/controller/util/node:go_default_library",
"//pkg/features:go_default_library",
"//pkg/kubelet/apis:go_default_library",
"//pkg/util/node:go_default_library",
"//pkg/util/taints:go_default_library",
"//staging/src/k8s.io/api/coordination/v1:go_default_library",
"//staging/src/k8s.io/api/core/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/api/equality:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/api/errors:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/apis/meta/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/labels:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/runtime:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/wait:go_default_library",
"//staging/src/k8s.io/apiserver/pkg/util/feature:go_default_library",
"//staging/src/k8s.io/client-go/informers/apps/v1:go_default_library",
"//staging/src/k8s.io/client-go/informers/coordination/v1:go_default_library",
"//staging/src/k8s.io/client-go/informers/core/v1:go_default_library",
"//staging/src/k8s.io/client-go/kubernetes:go_default_library",
"//staging/src/k8s.io/client-go/kubernetes/scheme:go_default_library",
"//staging/src/k8s.io/client-go/kubernetes/typed/core/v1:go_default_library",
"//staging/src/k8s.io/client-go/listers/apps/v1:go_default_library",
"//staging/src/k8s.io/client-go/listers/coordination/v1:go_default_library",
"//staging/src/k8s.io/client-go/listers/core/v1:go_default_library",
"//staging/src/k8s.io/client-go/tools/cache:go_default_library",
"//staging/src/k8s.io/client-go/tools/record:go_default_library",
"//staging/src/k8s.io/client-go/util/flowcontrol:go_default_library",
"//staging/src/k8s.io/client-go/util/workqueue:go_default_library",
"//staging/src/k8s.io/component-base/metrics:go_default_library",
"//staging/src/k8s.io/component-base/metrics/legacyregistry:go_default_library",
"//staging/src/k8s.io/component-base/metrics/prometheus/ratelimiter:go_default_library",
"//vendor/k8s.io/klog:go_default_library",
],
)
go_test(
name = "go_default_test",
srcs = ["node_lifecycle_controller_test.go"],
embed = [":go_default_library"],
deps = [
"//pkg/controller:go_default_library",
"//pkg/controller/nodelifecycle/scheduler:go_default_library",
"//pkg/controller/testutil:go_default_library",
"//pkg/controller/util/node:go_default_library",
"//pkg/features:go_default_library",
"//pkg/kubelet/apis:go_default_library",
"//pkg/util/node:go_default_library",
"//pkg/util/taints:go_default_library",
"//staging/src/k8s.io/api/apps/v1:go_default_library",
"//staging/src/k8s.io/api/coordination/v1:go_default_library",
"//staging/src/k8s.io/api/core/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/api/equality:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/api/resource:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/apis/meta/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/fields:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/labels:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/runtime:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/diff:go_default_library",
"//staging/src/k8s.io/apiserver/pkg/util/feature:go_default_library",
"//staging/src/k8s.io/client-go/informers:go_default_library",
"//staging/src/k8s.io/client-go/informers/apps/v1:go_default_library",
"//staging/src/k8s.io/client-go/informers/coordination/v1:go_default_library",
"//staging/src/k8s.io/client-go/informers/core/v1:go_default_library",
"//staging/src/k8s.io/client-go/kubernetes:go_default_library",
"//staging/src/k8s.io/client-go/kubernetes/fake:go_default_library",
"//staging/src/k8s.io/client-go/testing:go_default_library",
"//staging/src/k8s.io/component-base/featuregate/testing:go_default_library",
"//vendor/k8s.io/utils/pointer:go_default_library",
],
)
filegroup(
name = "package-srcs",
srcs = glob(["**"]),
tags = ["automanaged"],
visibility = ["//visibility:private"],
)
filegroup(
name = "all-srcs",
srcs = [
":package-srcs",
"//pkg/controller/nodelifecycle/config:all-srcs",
"//pkg/controller/nodelifecycle/scheduler:all-srcs",
],
tags = ["automanaged"],
visibility = ["//visibility:public"],
)

View File

@ -1,12 +0,0 @@
# See the OWNERS docs at https://go.k8s.io/owners
approvers:
- gmarek
- bowei
- k82cn
reviewers:
- gmarek
- smarterclayton
- ingvagabund
- aveshagarwal
- k82cn

View File

@ -1,83 +0,0 @@
/*
Copyright 2017 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 nodelifecycle
import (
"sync"
"k8s.io/component-base/metrics"
"k8s.io/component-base/metrics/legacyregistry"
)
const (
nodeControllerSubsystem = "node_collector"
zoneHealthStatisticKey = "zone_health"
zoneSizeKey = "zone_size"
zoneNoUnhealthyNodesKey = "unhealthy_nodes_in_zone"
evictionsNumberKey = "evictions_number"
)
var (
zoneHealth = metrics.NewGaugeVec(
&metrics.GaugeOpts{
Subsystem: nodeControllerSubsystem,
Name: zoneHealthStatisticKey,
Help: "Gauge measuring percentage of healthy nodes per zone.",
StabilityLevel: metrics.ALPHA,
},
[]string{"zone"},
)
zoneSize = metrics.NewGaugeVec(
&metrics.GaugeOpts{
Subsystem: nodeControllerSubsystem,
Name: zoneSizeKey,
Help: "Gauge measuring number of registered Nodes per zones.",
StabilityLevel: metrics.ALPHA,
},
[]string{"zone"},
)
unhealthyNodes = metrics.NewGaugeVec(
&metrics.GaugeOpts{
Subsystem: nodeControllerSubsystem,
Name: zoneNoUnhealthyNodesKey,
Help: "Gauge measuring number of not Ready Nodes per zones.",
StabilityLevel: metrics.ALPHA,
},
[]string{"zone"},
)
evictionsNumber = metrics.NewCounterVec(
&metrics.CounterOpts{
Subsystem: nodeControllerSubsystem,
Name: evictionsNumberKey,
Help: "Number of Node evictions that happened since current instance of NodeController started.",
StabilityLevel: metrics.ALPHA,
},
[]string{"zone"},
)
)
var registerMetrics sync.Once
// Register the metrics that are to be monitored.
func Register() {
registerMetrics.Do(func() {
legacyregistry.MustRegister(zoneHealth)
legacyregistry.MustRegister(zoneSize)
legacyregistry.MustRegister(unhealthyNodes)
legacyregistry.MustRegister(evictionsNumber)
})
}

File diff suppressed because it is too large Load Diff

View File

@ -1,64 +0,0 @@
load("@io_bazel_rules_go//go:def.bzl", "go_library", "go_test")
go_library(
name = "go_default_library",
srcs = [
"rate_limited_queue.go",
"taint_manager.go",
"timed_workers.go",
],
importpath = "k8s.io/kubernetes/pkg/controller/nodelifecycle/scheduler",
visibility = ["//visibility:public"],
deps = [
"//pkg/apis/core/helper:go_default_library",
"//pkg/apis/core/v1/helper:go_default_library",
"//staging/src/k8s.io/api/core/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/api/errors:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/apis/meta/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/types:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/runtime:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/sets:go_default_library",
"//staging/src/k8s.io/client-go/kubernetes:go_default_library",
"//staging/src/k8s.io/client-go/kubernetes/scheme:go_default_library",
"//staging/src/k8s.io/client-go/kubernetes/typed/core/v1:go_default_library",
"//staging/src/k8s.io/client-go/tools/record:go_default_library",
"//staging/src/k8s.io/client-go/util/flowcontrol:go_default_library",
"//staging/src/k8s.io/client-go/util/workqueue:go_default_library",
"//vendor/k8s.io/klog:go_default_library",
],
)
go_test(
name = "go_default_test",
srcs = [
"rate_limited_queue_test.go",
"taint_manager_test.go",
"timed_workers_test.go",
],
embed = [":go_default_library"],
deps = [
"//pkg/controller/testutil:go_default_library",
"//staging/src/k8s.io/api/core/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/apis/meta/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/fields:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/labels:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/sets:go_default_library",
"//staging/src/k8s.io/client-go/kubernetes/fake:go_default_library",
"//staging/src/k8s.io/client-go/testing:go_default_library",
"//staging/src/k8s.io/client-go/util/flowcontrol:go_default_library",
],
)
filegroup(
name = "package-srcs",
srcs = glob(["**"]),
tags = ["automanaged"],
visibility = ["//visibility:private"],
)
filegroup(
name = "all-srcs",
srcs = [":package-srcs"],
tags = ["automanaged"],
visibility = ["//visibility:public"],
)

View File

@ -1,309 +0,0 @@
/*
Copyright 2015 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 scheduler
import (
"container/heap"
"sync"
"time"
"k8s.io/apimachinery/pkg/util/sets"
"k8s.io/client-go/util/flowcontrol"
"k8s.io/klog"
)
const (
// NodeHealthUpdateRetry controls the number of retries of writing
// node health update.
NodeHealthUpdateRetry = 5
// NodeEvictionPeriod controls how often NodeController will try to
// evict Pods from non-responsive Nodes.
NodeEvictionPeriod = 100 * time.Millisecond
// EvictionRateLimiterBurst is the burst value for all eviction rate
// limiters
EvictionRateLimiterBurst = 1
)
// TimedValue is a value that should be processed at a designated time.
type TimedValue struct {
Value string
// UID could be anything that helps identify the value
UID interface{}
AddedAt time.Time
ProcessAt time.Time
}
// now is used to test time
var now = time.Now
// TimedQueue is a priority heap where the lowest ProcessAt is at the front of the queue
type TimedQueue []*TimedValue
// Len is the length of the queue.
func (h TimedQueue) Len() int { return len(h) }
// Less returns true if queue[i] < queue[j].
func (h TimedQueue) Less(i, j int) bool { return h[i].ProcessAt.Before(h[j].ProcessAt) }
// Swap swaps index i and j.
func (h TimedQueue) Swap(i, j int) { h[i], h[j] = h[j], h[i] }
// Push a new TimedValue on to the queue.
func (h *TimedQueue) Push(x interface{}) {
*h = append(*h, x.(*TimedValue))
}
// Pop the lowest ProcessAt item.
func (h *TimedQueue) Pop() interface{} {
old := *h
n := len(old)
x := old[n-1]
*h = old[0 : n-1]
return x
}
// UniqueQueue is a FIFO queue which additionally guarantees that any
// element can be added only once until it is removed.
type UniqueQueue struct {
lock sync.Mutex
queue TimedQueue
set sets.String
}
// Add a new value to the queue if it wasn't added before, or was
// explicitly removed by the Remove call. Returns true if new value
// was added.
func (q *UniqueQueue) Add(value TimedValue) bool {
q.lock.Lock()
defer q.lock.Unlock()
if q.set.Has(value.Value) {
return false
}
heap.Push(&q.queue, &value)
q.set.Insert(value.Value)
return true
}
// Replace replaces an existing value in the queue if it already
// exists, otherwise it does nothing. Returns true if the item was
// found.
func (q *UniqueQueue) Replace(value TimedValue) bool {
q.lock.Lock()
defer q.lock.Unlock()
for i := range q.queue {
if q.queue[i].Value != value.Value {
continue
}
heap.Remove(&q.queue, i)
heap.Push(&q.queue, &value)
return true
}
return false
}
// RemoveFromQueue the value from the queue, but keeps it in the set,
// so it won't be added second time. Returns true if something was
// removed.
func (q *UniqueQueue) RemoveFromQueue(value string) bool {
q.lock.Lock()
defer q.lock.Unlock()
if !q.set.Has(value) {
return false
}
for i, val := range q.queue {
if val.Value == value {
heap.Remove(&q.queue, i)
return true
}
}
return false
}
// Remove the value from the queue, so Get() call won't return it, and
// allow subsequent addition of the given value. If the value is not
// present does nothing and returns false.
func (q *UniqueQueue) Remove(value string) bool {
q.lock.Lock()
defer q.lock.Unlock()
if !q.set.Has(value) {
return false
}
q.set.Delete(value)
for i, val := range q.queue {
if val.Value == value {
heap.Remove(&q.queue, i)
return true
}
}
return true
}
// Get returns the oldest added value that wasn't returned yet.
func (q *UniqueQueue) Get() (TimedValue, bool) {
q.lock.Lock()
defer q.lock.Unlock()
if len(q.queue) == 0 {
return TimedValue{}, false
}
result := heap.Pop(&q.queue).(*TimedValue)
q.set.Delete(result.Value)
return *result, true
}
// Head returns the oldest added value that wasn't returned yet
// without removing it.
func (q *UniqueQueue) Head() (TimedValue, bool) {
q.lock.Lock()
defer q.lock.Unlock()
if len(q.queue) == 0 {
return TimedValue{}, false
}
result := q.queue[0]
return *result, true
}
// Clear removes all items from the queue and duplication preventing
// set.
func (q *UniqueQueue) Clear() {
q.lock.Lock()
defer q.lock.Unlock()
if q.queue.Len() > 0 {
q.queue = make(TimedQueue, 0)
}
if len(q.set) > 0 {
q.set = sets.NewString()
}
}
// RateLimitedTimedQueue is a unique item priority queue ordered by
// the expected next time of execution. It is also rate limited.
type RateLimitedTimedQueue struct {
queue UniqueQueue
limiterLock sync.Mutex
limiter flowcontrol.RateLimiter
}
// NewRateLimitedTimedQueue creates new queue which will use given
// RateLimiter to oversee execution.
func NewRateLimitedTimedQueue(limiter flowcontrol.RateLimiter) *RateLimitedTimedQueue {
return &RateLimitedTimedQueue{
queue: UniqueQueue{
queue: TimedQueue{},
set: sets.NewString(),
},
limiter: limiter,
}
}
// ActionFunc takes a timed value and returns false if the item must
// be retried, with an optional time.Duration if some minimum wait
// interval should be used.
type ActionFunc func(TimedValue) (bool, time.Duration)
// Try processes the queue.Ends prematurely if RateLimiter forbids an
// action and leak is true. Otherwise, requeues the item to be
// processed. Each value is processed once if fn returns true,
// otherwise it is added back to the queue. The returned remaining is
// used to identify the minimum time to execute the next item in the
// queue. The same value is processed only once unless Remove is
// explicitly called on it (it's done by the cancelPodEviction
// function in NodeController when Node becomes Ready again) TODO:
// figure out a good way to do garbage collection for all Nodes that
// were removed from the cluster.
func (q *RateLimitedTimedQueue) Try(fn ActionFunc) {
val, ok := q.queue.Head()
q.limiterLock.Lock()
defer q.limiterLock.Unlock()
for ok {
// rate limit the queue checking
if !q.limiter.TryAccept() {
klog.V(10).Infof("Try rate limited for value: %v", val)
// Try again later
break
}
now := now()
if now.Before(val.ProcessAt) {
break
}
if ok, wait := fn(val); !ok {
val.ProcessAt = now.Add(wait + 1)
q.queue.Replace(val)
} else {
q.queue.RemoveFromQueue(val.Value)
}
val, ok = q.queue.Head()
}
}
// Add value to the queue to be processed. Won't add the same
// value(comparison by value) a second time if it was already added
// and not removed.
func (q *RateLimitedTimedQueue) Add(value string, uid interface{}) bool {
now := now()
return q.queue.Add(TimedValue{
Value: value,
UID: uid,
AddedAt: now,
ProcessAt: now,
})
}
// Remove Node from the Evictor. The Node won't be processed until
// added again.
func (q *RateLimitedTimedQueue) Remove(value string) bool {
return q.queue.Remove(value)
}
// Clear removes all items from the queue
func (q *RateLimitedTimedQueue) Clear() {
q.queue.Clear()
}
// SwapLimiter safely swaps current limiter for this queue with the
// passed one if capacities or qps's differ.
func (q *RateLimitedTimedQueue) SwapLimiter(newQPS float32) {
q.limiterLock.Lock()
defer q.limiterLock.Unlock()
if q.limiter.QPS() == newQPS {
return
}
var newLimiter flowcontrol.RateLimiter
if newQPS <= 0 {
newLimiter = flowcontrol.NewFakeNeverRateLimiter()
} else {
newLimiter = flowcontrol.NewTokenBucketRateLimiter(newQPS, EvictionRateLimiterBurst)
// If we're currently waiting on limiter, we drain the new one - this is a good approach when Burst value is 1
// TODO: figure out if we need to support higher Burst values and decide on the drain logic, should we keep:
// - saturation (percentage of used tokens)
// - number of used tokens
// - number of available tokens
// - something else
if q.limiter.TryAccept() == false {
newLimiter.TryAccept()
}
}
q.limiter.Stop()
q.limiter = newLimiter
}

View File

@ -1,496 +0,0 @@
/*
Copyright 2017 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 scheduler
import (
"context"
"fmt"
"hash/fnv"
"io"
"math"
"sync"
"time"
"k8s.io/api/core/v1"
apierrors "k8s.io/apimachinery/pkg/api/errors"
metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
"k8s.io/apimachinery/pkg/types"
utilruntime "k8s.io/apimachinery/pkg/util/runtime"
clientset "k8s.io/client-go/kubernetes"
"k8s.io/client-go/kubernetes/scheme"
v1core "k8s.io/client-go/kubernetes/typed/core/v1"
"k8s.io/client-go/tools/record"
"k8s.io/client-go/util/workqueue"
"k8s.io/kubernetes/pkg/apis/core/helper"
v1helper "k8s.io/kubernetes/pkg/apis/core/v1/helper"
"k8s.io/klog"
)
const (
// TODO (k82cn): Figure out a reasonable number of workers/channels and propagate
// the number of workers up making it a parameter of Run() function.
// NodeUpdateChannelSize defines the size of channel for node update events.
NodeUpdateChannelSize = 10
// UpdateWorkerSize defines the size of workers for node update or/and pod update.
UpdateWorkerSize = 8
podUpdateChannelSize = 1
retries = 5
)
type nodeUpdateItem struct {
nodeName string
}
type podUpdateItem struct {
podName string
podNamespace string
nodeName string
}
func hash(val string, max int) int {
hasher := fnv.New32a()
io.WriteString(hasher, val)
return int(hasher.Sum32() % uint32(max))
}
// GetPodFunc returns the pod for the specified name/namespace, or a NotFound error if missing.
type GetPodFunc func(name, namespace string) (*v1.Pod, error)
// GetNodeFunc returns the node for the specified name, or a NotFound error if missing.
type GetNodeFunc func(name string) (*v1.Node, error)
// GetPodsByNodeNameFunc returns the list of pods assigned to the specified node.
type GetPodsByNodeNameFunc func(nodeName string) ([]*v1.Pod, error)
// NoExecuteTaintManager listens to Taint/Toleration changes and is responsible for removing Pods
// from Nodes tainted with NoExecute Taints.
type NoExecuteTaintManager struct {
client clientset.Interface
recorder record.EventRecorder
getPod GetPodFunc
getNode GetNodeFunc
getPodsAssignedToNode GetPodsByNodeNameFunc
taintEvictionQueue *TimedWorkerQueue
// keeps a map from nodeName to all noExecute taints on that Node
taintedNodesLock sync.Mutex
taintedNodes map[string][]v1.Taint
nodeUpdateChannels []chan nodeUpdateItem
podUpdateChannels []chan podUpdateItem
nodeUpdateQueue workqueue.Interface
podUpdateQueue workqueue.Interface
}
func deletePodHandler(c clientset.Interface, emitEventFunc func(types.NamespacedName)) func(args *WorkArgs) error {
return func(args *WorkArgs) error {
ns := args.NamespacedName.Namespace
name := args.NamespacedName.Name
klog.V(0).Infof("NoExecuteTaintManager is deleting Pod: %v", args.NamespacedName.String())
if emitEventFunc != nil {
emitEventFunc(args.NamespacedName)
}
var err error
for i := 0; i < retries; i++ {
err = c.CoreV1().Pods(ns).Delete(context.TODO(), name, metav1.DeleteOptions{})
if err == nil {
break
}
time.Sleep(10 * time.Millisecond)
}
return err
}
}
func getNoExecuteTaints(taints []v1.Taint) []v1.Taint {
result := []v1.Taint{}
for i := range taints {
if taints[i].Effect == v1.TaintEffectNoExecute {
result = append(result, taints[i])
}
}
return result
}
// getMinTolerationTime returns minimal toleration time from the given slice, or -1 if it's infinite.
func getMinTolerationTime(tolerations []v1.Toleration) time.Duration {
minTolerationTime := int64(math.MaxInt64)
if len(tolerations) == 0 {
return 0
}
for i := range tolerations {
if tolerations[i].TolerationSeconds != nil {
tolerationSeconds := *(tolerations[i].TolerationSeconds)
if tolerationSeconds <= 0 {
return 0
} else if tolerationSeconds < minTolerationTime {
minTolerationTime = tolerationSeconds
}
}
}
if minTolerationTime == int64(math.MaxInt64) {
return -1
}
return time.Duration(minTolerationTime) * time.Second
}
// NewNoExecuteTaintManager creates a new NoExecuteTaintManager that will use passed clientset to
// communicate with the API server.
func NewNoExecuteTaintManager(c clientset.Interface, getPod GetPodFunc, getNode GetNodeFunc, getPodsAssignedToNode GetPodsByNodeNameFunc) *NoExecuteTaintManager {
eventBroadcaster := record.NewBroadcaster()
recorder := eventBroadcaster.NewRecorder(scheme.Scheme, v1.EventSource{Component: "taint-controller"})
eventBroadcaster.StartLogging(klog.Infof)
if c != nil {
klog.V(0).Infof("Sending events to api server.")
eventBroadcaster.StartRecordingToSink(&v1core.EventSinkImpl{Interface: c.CoreV1().Events("")})
} else {
klog.Fatalf("kubeClient is nil when starting NodeController")
}
tm := &NoExecuteTaintManager{
client: c,
recorder: recorder,
getPod: getPod,
getNode: getNode,
getPodsAssignedToNode: getPodsAssignedToNode,
taintedNodes: make(map[string][]v1.Taint),
nodeUpdateQueue: workqueue.NewNamed("noexec_taint_node"),
podUpdateQueue: workqueue.NewNamed("noexec_taint_pod"),
}
tm.taintEvictionQueue = CreateWorkerQueue(deletePodHandler(c, tm.emitPodDeletionEvent))
return tm
}
// Run starts NoExecuteTaintManager which will run in loop until `stopCh` is closed.
func (tc *NoExecuteTaintManager) Run(stopCh <-chan struct{}) {
klog.V(0).Infof("Starting NoExecuteTaintManager")
for i := 0; i < UpdateWorkerSize; i++ {
tc.nodeUpdateChannels = append(tc.nodeUpdateChannels, make(chan nodeUpdateItem, NodeUpdateChannelSize))
tc.podUpdateChannels = append(tc.podUpdateChannels, make(chan podUpdateItem, podUpdateChannelSize))
}
// Functions that are responsible for taking work items out of the workqueues and putting them
// into channels.
go func(stopCh <-chan struct{}) {
for {
item, shutdown := tc.nodeUpdateQueue.Get()
if shutdown {
break
}
nodeUpdate := item.(nodeUpdateItem)
hash := hash(nodeUpdate.nodeName, UpdateWorkerSize)
select {
case <-stopCh:
tc.nodeUpdateQueue.Done(item)
return
case tc.nodeUpdateChannels[hash] <- nodeUpdate:
// tc.nodeUpdateQueue.Done is called by the nodeUpdateChannels worker
}
}
}(stopCh)
go func(stopCh <-chan struct{}) {
for {
item, shutdown := tc.podUpdateQueue.Get()
if shutdown {
break
}
// The fact that pods are processed by the same worker as nodes is used to avoid races
// between node worker setting tc.taintedNodes and pod worker reading this to decide
// whether to delete pod.
// It's possible that even without this assumption this code is still correct.
podUpdate := item.(podUpdateItem)
hash := hash(podUpdate.nodeName, UpdateWorkerSize)
select {
case <-stopCh:
tc.podUpdateQueue.Done(item)
return
case tc.podUpdateChannels[hash] <- podUpdate:
// tc.podUpdateQueue.Done is called by the podUpdateChannels worker
}
}
}(stopCh)
wg := sync.WaitGroup{}
wg.Add(UpdateWorkerSize)
for i := 0; i < UpdateWorkerSize; i++ {
go tc.worker(i, wg.Done, stopCh)
}
wg.Wait()
}
func (tc *NoExecuteTaintManager) worker(worker int, done func(), stopCh <-chan struct{}) {
defer done()
// When processing events we want to prioritize Node updates over Pod updates,
// as NodeUpdates that interest NoExecuteTaintManager should be handled as soon as possible -
// we don't want user (or system) to wait until PodUpdate queue is drained before it can
// start evicting Pods from tainted Nodes.
for {
select {
case <-stopCh:
return
case nodeUpdate := <-tc.nodeUpdateChannels[worker]:
tc.handleNodeUpdate(nodeUpdate)
tc.nodeUpdateQueue.Done(nodeUpdate)
case podUpdate := <-tc.podUpdateChannels[worker]:
// If we found a Pod update we need to empty Node queue first.
priority:
for {
select {
case nodeUpdate := <-tc.nodeUpdateChannels[worker]:
tc.handleNodeUpdate(nodeUpdate)
tc.nodeUpdateQueue.Done(nodeUpdate)
default:
break priority
}
}
// After Node queue is emptied we process podUpdate.
tc.handlePodUpdate(podUpdate)
tc.podUpdateQueue.Done(podUpdate)
}
}
}
// PodUpdated is used to notify NoExecuteTaintManager about Pod changes.
func (tc *NoExecuteTaintManager) PodUpdated(oldPod *v1.Pod, newPod *v1.Pod) {
podName := ""
podNamespace := ""
nodeName := ""
oldTolerations := []v1.Toleration{}
if oldPod != nil {
podName = oldPod.Name
podNamespace = oldPod.Namespace
nodeName = oldPod.Spec.NodeName
oldTolerations = oldPod.Spec.Tolerations
}
newTolerations := []v1.Toleration{}
if newPod != nil {
podName = newPod.Name
podNamespace = newPod.Namespace
nodeName = newPod.Spec.NodeName
newTolerations = newPod.Spec.Tolerations
}
if oldPod != nil && newPod != nil && helper.Semantic.DeepEqual(oldTolerations, newTolerations) && oldPod.Spec.NodeName == newPod.Spec.NodeName {
return
}
updateItem := podUpdateItem{
podName: podName,
podNamespace: podNamespace,
nodeName: nodeName,
}
tc.podUpdateQueue.Add(updateItem)
}
// NodeUpdated is used to notify NoExecuteTaintManager about Node changes.
func (tc *NoExecuteTaintManager) NodeUpdated(oldNode *v1.Node, newNode *v1.Node) {
nodeName := ""
oldTaints := []v1.Taint{}
if oldNode != nil {
nodeName = oldNode.Name
oldTaints = getNoExecuteTaints(oldNode.Spec.Taints)
}
newTaints := []v1.Taint{}
if newNode != nil {
nodeName = newNode.Name
newTaints = getNoExecuteTaints(newNode.Spec.Taints)
}
if oldNode != nil && newNode != nil && helper.Semantic.DeepEqual(oldTaints, newTaints) {
return
}
updateItem := nodeUpdateItem{
nodeName: nodeName,
}
tc.nodeUpdateQueue.Add(updateItem)
}
func (tc *NoExecuteTaintManager) cancelWorkWithEvent(nsName types.NamespacedName) {
if tc.taintEvictionQueue.CancelWork(nsName.String()) {
tc.emitCancelPodDeletionEvent(nsName)
}
}
func (tc *NoExecuteTaintManager) processPodOnNode(
podNamespacedName types.NamespacedName,
nodeName string,
tolerations []v1.Toleration,
taints []v1.Taint,
now time.Time,
) {
if len(taints) == 0 {
tc.cancelWorkWithEvent(podNamespacedName)
}
allTolerated, usedTolerations := v1helper.GetMatchingTolerations(taints, tolerations)
if !allTolerated {
klog.V(2).Infof("Not all taints are tolerated after update for Pod %v on %v", podNamespacedName.String(), nodeName)
// We're canceling scheduled work (if any), as we're going to delete the Pod right away.
tc.cancelWorkWithEvent(podNamespacedName)
tc.taintEvictionQueue.AddWork(NewWorkArgs(podNamespacedName.Name, podNamespacedName.Namespace), time.Now(), time.Now())
return
}
minTolerationTime := getMinTolerationTime(usedTolerations)
// getMinTolerationTime returns negative value to denote infinite toleration.
if minTolerationTime < 0 {
klog.V(4).Infof("New tolerations for %v tolerate forever. Scheduled deletion won't be cancelled if already scheduled.", podNamespacedName.String())
return
}
startTime := now
triggerTime := startTime.Add(minTolerationTime)
scheduledEviction := tc.taintEvictionQueue.GetWorkerUnsafe(podNamespacedName.String())
if scheduledEviction != nil {
startTime = scheduledEviction.CreatedAt
if startTime.Add(minTolerationTime).Before(triggerTime) {
return
}
tc.cancelWorkWithEvent(podNamespacedName)
}
tc.taintEvictionQueue.AddWork(NewWorkArgs(podNamespacedName.Name, podNamespacedName.Namespace), startTime, triggerTime)
}
func (tc *NoExecuteTaintManager) handlePodUpdate(podUpdate podUpdateItem) {
pod, err := tc.getPod(podUpdate.podName, podUpdate.podNamespace)
if err != nil {
if apierrors.IsNotFound(err) {
// Delete
podNamespacedName := types.NamespacedName{Namespace: podUpdate.podNamespace, Name: podUpdate.podName}
klog.V(4).Infof("Noticed pod deletion: %#v", podNamespacedName)
tc.cancelWorkWithEvent(podNamespacedName)
return
}
utilruntime.HandleError(fmt.Errorf("could not get pod %s/%s: %v", podUpdate.podName, podUpdate.podNamespace, err))
return
}
// We key the workqueue and shard workers by nodeName. If we don't match the current state we should not be the one processing the current object.
if pod.Spec.NodeName != podUpdate.nodeName {
return
}
// Create or Update
podNamespacedName := types.NamespacedName{Namespace: pod.Namespace, Name: pod.Name}
klog.V(4).Infof("Noticed pod update: %#v", podNamespacedName)
nodeName := pod.Spec.NodeName
if nodeName == "" {
return
}
taints, ok := func() ([]v1.Taint, bool) {
tc.taintedNodesLock.Lock()
defer tc.taintedNodesLock.Unlock()
taints, ok := tc.taintedNodes[nodeName]
return taints, ok
}()
// It's possible that Node was deleted, or Taints were removed before, which triggered
// eviction cancelling if it was needed.
if !ok {
return
}
tc.processPodOnNode(podNamespacedName, nodeName, pod.Spec.Tolerations, taints, time.Now())
}
func (tc *NoExecuteTaintManager) handleNodeUpdate(nodeUpdate nodeUpdateItem) {
node, err := tc.getNode(nodeUpdate.nodeName)
if err != nil {
if apierrors.IsNotFound(err) {
// Delete
klog.V(4).Infof("Noticed node deletion: %#v", nodeUpdate.nodeName)
tc.taintedNodesLock.Lock()
defer tc.taintedNodesLock.Unlock()
delete(tc.taintedNodes, nodeUpdate.nodeName)
return
}
utilruntime.HandleError(fmt.Errorf("cannot get node %s: %v", nodeUpdate.nodeName, err))
return
}
// Create or Update
klog.V(4).Infof("Noticed node update: %#v", nodeUpdate)
taints := getNoExecuteTaints(node.Spec.Taints)
func() {
tc.taintedNodesLock.Lock()
defer tc.taintedNodesLock.Unlock()
klog.V(4).Infof("Updating known taints on node %v: %v", node.Name, taints)
if len(taints) == 0 {
delete(tc.taintedNodes, node.Name)
} else {
tc.taintedNodes[node.Name] = taints
}
}()
// This is critical that we update tc.taintedNodes before we call getPodsAssignedToNode:
// getPodsAssignedToNode can be delayed as long as all future updates to pods will call
// tc.PodUpdated which will use tc.taintedNodes to potentially delete delayed pods.
pods, err := tc.getPodsAssignedToNode(node.Name)
if err != nil {
klog.Errorf(err.Error())
return
}
if len(pods) == 0 {
return
}
// Short circuit, to make this controller a bit faster.
if len(taints) == 0 {
klog.V(4).Infof("All taints were removed from the Node %v. Cancelling all evictions...", node.Name)
for i := range pods {
tc.cancelWorkWithEvent(types.NamespacedName{Namespace: pods[i].Namespace, Name: pods[i].Name})
}
return
}
now := time.Now()
for _, pod := range pods {
podNamespacedName := types.NamespacedName{Namespace: pod.Namespace, Name: pod.Name}
tc.processPodOnNode(podNamespacedName, node.Name, pod.Spec.Tolerations, taints, now)
}
}
func (tc *NoExecuteTaintManager) emitPodDeletionEvent(nsName types.NamespacedName) {
if tc.recorder == nil {
return
}
ref := &v1.ObjectReference{
Kind: "Pod",
Name: nsName.Name,
Namespace: nsName.Namespace,
}
tc.recorder.Eventf(ref, v1.EventTypeNormal, "TaintManagerEviction", "Marking for deletion Pod %s", nsName.String())
}
func (tc *NoExecuteTaintManager) emitCancelPodDeletionEvent(nsName types.NamespacedName) {
if tc.recorder == nil {
return
}
ref := &v1.ObjectReference{
Kind: "Pod",
Name: nsName.Name,
Namespace: nsName.Namespace,
}
tc.recorder.Eventf(ref, v1.EventTypeNormal, "TaintManagerEviction", "Cancelling deletion of Pod %s", nsName.String())
}

View File

@ -1,145 +0,0 @@
/*
Copyright 2015 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 scheduler
import (
"sync"
"time"
"k8s.io/apimachinery/pkg/types"
"k8s.io/klog"
)
// WorkArgs keeps arguments that will be passed to the function executed by the worker.
type WorkArgs struct {
NamespacedName types.NamespacedName
}
// KeyFromWorkArgs creates a key for the given `WorkArgs`
func (w *WorkArgs) KeyFromWorkArgs() string {
return w.NamespacedName.String()
}
// NewWorkArgs is a helper function to create new `WorkArgs`
func NewWorkArgs(name, namespace string) *WorkArgs {
return &WorkArgs{types.NamespacedName{Namespace: namespace, Name: name}}
}
// TimedWorker is a responsible for executing a function no earlier than at FireAt time.
type TimedWorker struct {
WorkItem *WorkArgs
CreatedAt time.Time
FireAt time.Time
Timer *time.Timer
}
// CreateWorker creates a TimedWorker that will execute `f` not earlier than `fireAt`.
func CreateWorker(args *WorkArgs, createdAt time.Time, fireAt time.Time, f func(args *WorkArgs) error) *TimedWorker {
delay := fireAt.Sub(createdAt)
if delay <= 0 {
go f(args)
return nil
}
timer := time.AfterFunc(delay, func() { f(args) })
return &TimedWorker{
WorkItem: args,
CreatedAt: createdAt,
FireAt: fireAt,
Timer: timer,
}
}
// Cancel cancels the execution of function by the `TimedWorker`
func (w *TimedWorker) Cancel() {
if w != nil {
w.Timer.Stop()
}
}
// TimedWorkerQueue keeps a set of TimedWorkers that are still wait for execution.
type TimedWorkerQueue struct {
sync.Mutex
// map of workers keyed by string returned by 'KeyFromWorkArgs' from the given worker.
workers map[string]*TimedWorker
workFunc func(args *WorkArgs) error
}
// CreateWorkerQueue creates a new TimedWorkerQueue for workers that will execute
// given function `f`.
func CreateWorkerQueue(f func(args *WorkArgs) error) *TimedWorkerQueue {
return &TimedWorkerQueue{
workers: make(map[string]*TimedWorker),
workFunc: f,
}
}
func (q *TimedWorkerQueue) getWrappedWorkerFunc(key string) func(args *WorkArgs) error {
return func(args *WorkArgs) error {
err := q.workFunc(args)
q.Lock()
defer q.Unlock()
if err == nil {
// To avoid duplicated calls we keep the key in the queue, to prevent
// subsequent additions.
q.workers[key] = nil
} else {
delete(q.workers, key)
}
return err
}
}
// AddWork adds a work to the WorkerQueue which will be executed not earlier than `fireAt`.
func (q *TimedWorkerQueue) AddWork(args *WorkArgs, createdAt time.Time, fireAt time.Time) {
key := args.KeyFromWorkArgs()
klog.V(4).Infof("Adding TimedWorkerQueue item %v at %v to be fired at %v", key, createdAt, fireAt)
q.Lock()
defer q.Unlock()
if _, exists := q.workers[key]; exists {
klog.Warningf("Trying to add already existing work for %+v. Skipping.", args)
return
}
worker := CreateWorker(args, createdAt, fireAt, q.getWrappedWorkerFunc(key))
q.workers[key] = worker
}
// CancelWork removes scheduled function execution from the queue. Returns true if work was cancelled.
func (q *TimedWorkerQueue) CancelWork(key string) bool {
q.Lock()
defer q.Unlock()
worker, found := q.workers[key]
result := false
if found {
klog.V(4).Infof("Cancelling TimedWorkerQueue item %v at %v", key, time.Now())
if worker != nil {
result = true
worker.Cancel()
}
delete(q.workers, key)
}
return result
}
// GetWorkerUnsafe returns a TimedWorker corresponding to the given key.
// Unsafe method - workers have attached goroutines which can fire after this function is called.
func (q *TimedWorkerQueue) GetWorkerUnsafe(key string) *TimedWorker {
q.Lock()
defer q.Unlock()
return q.workers[key]
}

View File

@ -1,39 +0,0 @@
load("@io_bazel_rules_go//go:def.bzl", "go_library")
go_library(
name = "go_default_library",
srcs = ["controller_utils.go"],
importpath = "k8s.io/kubernetes/pkg/controller/util/node",
visibility = ["//visibility:public"],
deps = [
"//pkg/api/v1/pod:go_default_library",
"//pkg/controller:go_default_library",
"//pkg/kubelet/util/format:go_default_library",
"//pkg/util/node:go_default_library",
"//staging/src/k8s.io/api/core/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/api/errors:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/apis/meta/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/types:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/errors:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/runtime:go_default_library",
"//staging/src/k8s.io/client-go/kubernetes:go_default_library",
"//staging/src/k8s.io/client-go/listers/apps/v1:go_default_library",
"//staging/src/k8s.io/client-go/tools/cache:go_default_library",
"//staging/src/k8s.io/client-go/tools/record:go_default_library",
"//vendor/k8s.io/klog:go_default_library",
],
)
filegroup(
name = "package-srcs",
srcs = glob(["**"]),
tags = ["automanaged"],
visibility = ["//visibility:private"],
)
filegroup(
name = "all-srcs",
srcs = [":package-srcs"],
tags = ["automanaged"],
visibility = ["//visibility:public"],
)

View File

@ -1,299 +0,0 @@
/*
Copyright 2016 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 node
import (
"context"
"fmt"
"strings"
apierrors "k8s.io/apimachinery/pkg/api/errors"
metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
"k8s.io/apimachinery/pkg/types"
utilerrors "k8s.io/apimachinery/pkg/util/errors"
utilruntime "k8s.io/apimachinery/pkg/util/runtime"
"k8s.io/client-go/tools/cache"
"k8s.io/client-go/tools/record"
"k8s.io/api/core/v1"
clientset "k8s.io/client-go/kubernetes"
appsv1listers "k8s.io/client-go/listers/apps/v1"
utilpod "k8s.io/kubernetes/pkg/api/v1/pod"
"k8s.io/kubernetes/pkg/controller"
"k8s.io/kubernetes/pkg/kubelet/util/format"
nodepkg "k8s.io/kubernetes/pkg/util/node"
"k8s.io/klog"
)
// DeletePods will delete all pods from master running on given node,
// and return true if any pods were deleted, or were found pending
// deletion.
func DeletePods(kubeClient clientset.Interface, pods []*v1.Pod, recorder record.EventRecorder, nodeName, nodeUID string, daemonStore appsv1listers.DaemonSetLister) (bool, error) {
remaining := false
var updateErrList []error
if len(pods) > 0 {
RecordNodeEvent(recorder, nodeName, nodeUID, v1.EventTypeNormal, "DeletingAllPods", fmt.Sprintf("Deleting all Pods from Node %v.", nodeName))
}
for i := range pods {
// Defensive check, also needed for tests.
if pods[i].Spec.NodeName != nodeName {
continue
}
// Pod will be modified, so making copy is required.
pod := pods[i].DeepCopy()
// Set reason and message in the pod object.
if _, err := SetPodTerminationReason(kubeClient, pod, nodeName); err != nil {
if apierrors.IsConflict(err) {
updateErrList = append(updateErrList,
fmt.Errorf("update status failed for pod %q: %v", format.Pod(pod), err))
continue
}
}
// if the pod has already been marked for deletion, we still return true that there are remaining pods.
if pod.DeletionGracePeriodSeconds != nil {
remaining = true
continue
}
// if the pod is managed by a daemonset, ignore it
if _, err := daemonStore.GetPodDaemonSets(pod); err == nil {
// No error means at least one daemonset was found
continue
}
klog.V(2).Infof("Starting deletion of pod %v/%v", pod.Namespace, pod.Name)
recorder.Eventf(pod, v1.EventTypeNormal, "NodeControllerEviction", "Marking for deletion Pod %s from Node %s", pod.Name, nodeName)
if err := kubeClient.CoreV1().Pods(pod.Namespace).Delete(context.TODO(), pod.Name, metav1.DeleteOptions{}); err != nil {
if apierrors.IsNotFound(err) {
// NotFound error means that pod was already deleted.
// There is nothing left to do with this pod.
continue
}
return false, err
}
remaining = true
}
if len(updateErrList) > 0 {
return false, utilerrors.NewAggregate(updateErrList)
}
return remaining, nil
}
// SetPodTerminationReason attempts to set a reason and message in the
// pod status, updates it in the apiserver, and returns an error if it
// encounters one.
func SetPodTerminationReason(kubeClient clientset.Interface, pod *v1.Pod, nodeName string) (*v1.Pod, error) {
if pod.Status.Reason == nodepkg.NodeUnreachablePodReason {
return pod, nil
}
pod.Status.Reason = nodepkg.NodeUnreachablePodReason
pod.Status.Message = fmt.Sprintf(nodepkg.NodeUnreachablePodMessage, nodeName, pod.Name)
var updatedPod *v1.Pod
var err error
if updatedPod, err = kubeClient.CoreV1().Pods(pod.Namespace).UpdateStatus(context.TODO(), pod, metav1.UpdateOptions{}); err != nil {
return nil, err
}
return updatedPod, nil
}
// MarkPodsNotReady updates ready status of given pods running on
// given node from master return true if success
func MarkPodsNotReady(kubeClient clientset.Interface, pods []*v1.Pod, nodeName string) error {
klog.V(2).Infof("Update ready status of pods on node [%v]", nodeName)
errMsg := []string{}
for i := range pods {
// Defensive check, also needed for tests.
if pods[i].Spec.NodeName != nodeName {
continue
}
// Pod will be modified, so making copy is required.
pod := pods[i].DeepCopy()
for _, cond := range pod.Status.Conditions {
if cond.Type == v1.PodReady {
cond.Status = v1.ConditionFalse
if !utilpod.UpdatePodCondition(&pod.Status, &cond) {
break
}
klog.V(2).Infof("Updating ready status of pod %v to false", pod.Name)
_, err := kubeClient.CoreV1().Pods(pod.Namespace).UpdateStatus(context.TODO(), pod, metav1.UpdateOptions{})
if err != nil {
if apierrors.IsNotFound(err) {
// NotFound error means that pod was already deleted.
// There is nothing left to do with this pod.
continue
}
klog.Warningf("Failed to update status for pod %q: %v", format.Pod(pod), err)
errMsg = append(errMsg, fmt.Sprintf("%v", err))
}
break
}
}
}
if len(errMsg) == 0 {
return nil
}
return fmt.Errorf("%v", strings.Join(errMsg, "; "))
}
// RecordNodeEvent records a event related to a node.
func RecordNodeEvent(recorder record.EventRecorder, nodeName, nodeUID, eventtype, reason, event string) {
ref := &v1.ObjectReference{
APIVersion: "v1",
Kind: "Node",
Name: nodeName,
UID: types.UID(nodeUID),
Namespace: "",
}
klog.V(2).Infof("Recording %s event message for node %s", event, nodeName)
recorder.Eventf(ref, eventtype, reason, "Node %s event: %s", nodeName, event)
}
// RecordNodeStatusChange records a event related to a node status change. (Common to lifecycle and ipam)
func RecordNodeStatusChange(recorder record.EventRecorder, node *v1.Node, newStatus string) {
ref := &v1.ObjectReference{
APIVersion: "v1",
Kind: "Node",
Name: node.Name,
UID: node.UID,
Namespace: "",
}
klog.V(2).Infof("Recording status change %s event message for node %s", newStatus, node.Name)
// TODO: This requires a transaction, either both node status is updated
// and event is recorded or neither should happen, see issue #6055.
recorder.Eventf(ref, v1.EventTypeNormal, newStatus, "Node %s status is now: %s", node.Name, newStatus)
}
// SwapNodeControllerTaint returns true in case of success and false
// otherwise.
func SwapNodeControllerTaint(kubeClient clientset.Interface, taintsToAdd, taintsToRemove []*v1.Taint, node *v1.Node) bool {
for _, taintToAdd := range taintsToAdd {
now := metav1.Now()
taintToAdd.TimeAdded = &now
}
err := controller.AddOrUpdateTaintOnNode(kubeClient, node.Name, taintsToAdd...)
if err != nil {
utilruntime.HandleError(
fmt.Errorf(
"unable to taint %+v unresponsive Node %q: %v",
taintsToAdd,
node.Name,
err))
return false
}
klog.V(4).Infof("Added %+v Taint to Node %v", taintsToAdd, node.Name)
err = controller.RemoveTaintOffNode(kubeClient, node.Name, node, taintsToRemove...)
if err != nil {
utilruntime.HandleError(
fmt.Errorf(
"unable to remove %+v unneeded taint from unresponsive Node %q: %v",
taintsToRemove,
node.Name,
err))
return false
}
klog.V(4).Infof("Made sure that Node %+v has no %v Taint", node.Name, taintsToRemove)
return true
}
// AddOrUpdateLabelsOnNode updates the labels on the node and returns true on
// success and false on failure.
func AddOrUpdateLabelsOnNode(kubeClient clientset.Interface, labelsToUpdate map[string]string, node *v1.Node) bool {
err := controller.AddOrUpdateLabelsOnNode(kubeClient, node.Name, labelsToUpdate)
if err != nil {
utilruntime.HandleError(
fmt.Errorf(
"unable to update labels %+v for Node %q: %v",
labelsToUpdate,
node.Name,
err))
return false
}
klog.V(4).Infof("Updated labels %+v to Node %v", labelsToUpdate, node.Name)
return true
}
// CreateAddNodeHandler creates an add node handler.
func CreateAddNodeHandler(f func(node *v1.Node) error) func(obj interface{}) {
return func(originalObj interface{}) {
node := originalObj.(*v1.Node).DeepCopy()
if err := f(node); err != nil {
utilruntime.HandleError(fmt.Errorf("Error while processing Node Add: %v", err))
}
}
}
// CreateUpdateNodeHandler creates a node update handler. (Common to lifecycle and ipam)
func CreateUpdateNodeHandler(f func(oldNode, newNode *v1.Node) error) func(oldObj, newObj interface{}) {
return func(origOldObj, origNewObj interface{}) {
node := origNewObj.(*v1.Node).DeepCopy()
prevNode := origOldObj.(*v1.Node).DeepCopy()
if err := f(prevNode, node); err != nil {
utilruntime.HandleError(fmt.Errorf("Error while processing Node Add/Delete: %v", err))
}
}
}
// CreateDeleteNodeHandler creates a delete node handler. (Common to lifecycle and ipam)
func CreateDeleteNodeHandler(f func(node *v1.Node) error) func(obj interface{}) {
return func(originalObj interface{}) {
originalNode, isNode := originalObj.(*v1.Node)
// We can get DeletedFinalStateUnknown instead of *v1.Node here and
// we need to handle that correctly. #34692
if !isNode {
deletedState, ok := originalObj.(cache.DeletedFinalStateUnknown)
if !ok {
klog.Errorf("Received unexpected object: %v", originalObj)
return
}
originalNode, ok = deletedState.Obj.(*v1.Node)
if !ok {
klog.Errorf("DeletedFinalStateUnknown contained non-Node object: %v", deletedState.Obj)
return
}
}
node := originalNode.DeepCopy()
if err := f(node); err != nil {
utilruntime.HandleError(fmt.Errorf("Error while processing Node Add/Delete: %v", err))
}
}
}
// GetNodeCondition extracts the provided condition from the given status and returns that.
// Returns nil and -1 if the condition is not present, and the index of the located condition.
func GetNodeCondition(status *v1.NodeStatus, conditionType v1.NodeConditionType) (int, *v1.NodeCondition) {
if status == nil {
return -1, nil
}
for i := range status.Conditions {
if status.Conditions[i].Type == conditionType {
return i, &status.Conditions[i]
}
}
return -1, nil
}

View File

@ -1,37 +0,0 @@
load("@io_bazel_rules_go//go:def.bzl", "go_library")
go_library(
name = "go_default_library",
srcs = ["util.go"],
importpath = "k8s.io/kubernetes/pkg/controller/volume/persistentvolume/util",
visibility = ["//visibility:public"],
deps = [
"//pkg/apis/core/v1/helper:go_default_library",
"//pkg/features:go_default_library",
"//pkg/volume/util:go_default_library",
"//staging/src/k8s.io/api/core/v1:go_default_library",
"//staging/src/k8s.io/api/storage/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/api/errors:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/api/resource:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/apis/meta/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/labels:go_default_library",
"//staging/src/k8s.io/apiserver/pkg/util/feature:go_default_library",
"//staging/src/k8s.io/client-go/kubernetes/scheme:go_default_library",
"//staging/src/k8s.io/client-go/listers/storage/v1:go_default_library",
"//staging/src/k8s.io/client-go/tools/reference:go_default_library",
],
)
filegroup(
name = "package-srcs",
srcs = glob(["**"]),
tags = ["automanaged"],
visibility = ["//visibility:private"],
)
filegroup(
name = "all-srcs",
srcs = [":package-srcs"],
tags = ["automanaged"],
visibility = ["//visibility:public"],
)

View File

@ -1,360 +0,0 @@
/*
Copyright 2019 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 persistentvolume
import (
"fmt"
v1 "k8s.io/api/core/v1"
storage "k8s.io/api/storage/v1"
apierrors "k8s.io/apimachinery/pkg/api/errors"
"k8s.io/apimachinery/pkg/api/resource"
metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
"k8s.io/apimachinery/pkg/labels"
utilfeature "k8s.io/apiserver/pkg/util/feature"
"k8s.io/client-go/kubernetes/scheme"
storagelisters "k8s.io/client-go/listers/storage/v1"
"k8s.io/client-go/tools/reference"
v1helper "k8s.io/kubernetes/pkg/apis/core/v1/helper"
"k8s.io/kubernetes/pkg/features"
volumeutil "k8s.io/kubernetes/pkg/volume/util"
)
const (
// AnnBindCompleted Annotation applies to PVCs. It indicates that the lifecycle
// of the PVC has passed through the initial setup. This information changes how
// we interpret some observations of the state of the objects. Value of this
// Annotation does not matter.
AnnBindCompleted = "pv.kubernetes.io/bind-completed"
// AnnBoundByController annotation applies to PVs and PVCs. It indicates that
// the binding (PV->PVC or PVC->PV) was installed by the controller. The
// absence of this annotation means the binding was done by the user (i.e.
// pre-bound). Value of this annotation does not matter.
// External PV binders must bind PV the same way as PV controller, otherwise PV
// controller may not handle it correctly.
AnnBoundByController = "pv.kubernetes.io/bound-by-controller"
// AnnSelectedNode annotation is added to a PVC that has been triggered by scheduler to
// be dynamically provisioned. Its value is the name of the selected node.
AnnSelectedNode = "volume.kubernetes.io/selected-node"
// NotSupportedProvisioner is a special provisioner name which can be set
// in storage class to indicate dynamic provisioning is not supported by
// the storage.
NotSupportedProvisioner = "kubernetes.io/no-provisioner"
// AnnDynamicallyProvisioned annotation is added to a PV that has been dynamically provisioned by
// Kubernetes. Its value is name of volume plugin that created the volume.
// It serves both user (to show where a PV comes from) and Kubernetes (to
// recognize dynamically provisioned PVs in its decisions).
AnnDynamicallyProvisioned = "pv.kubernetes.io/provisioned-by"
// AnnMigratedTo annotation is added to a PVC and PV that is supposed to be
// dynamically provisioned/deleted by by its corresponding CSI driver
// through the CSIMigration feature flags. When this annotation is set the
// Kubernetes components will "stand-down" and the external-provisioner will
// act on the objects
AnnMigratedTo = "pv.kubernetes.io/migrated-to"
// AnnStorageProvisioner annotation is added to a PVC that is supposed to be dynamically
// provisioned. Its value is name of volume plugin that is supposed to provision
// a volume for this PVC.
AnnStorageProvisioner = "volume.beta.kubernetes.io/storage-provisioner"
)
// IsDelayBindingProvisioning checks if claim provisioning with selected-node annotation
func IsDelayBindingProvisioning(claim *v1.PersistentVolumeClaim) bool {
// When feature VolumeScheduling enabled,
// Scheduler signal to the PV controller to start dynamic
// provisioning by setting the "AnnSelectedNode" annotation
// in the PVC
_, ok := claim.Annotations[AnnSelectedNode]
return ok
}
// IsDelayBindingMode checks if claim is in delay binding mode.
func IsDelayBindingMode(claim *v1.PersistentVolumeClaim, classLister storagelisters.StorageClassLister) (bool, error) {
className := v1helper.GetPersistentVolumeClaimClass(claim)
if className == "" {
return false, nil
}
class, err := classLister.Get(className)
if err != nil {
if apierrors.IsNotFound(err) {
return false, nil
}
return false, err
}
if class.VolumeBindingMode == nil {
return false, fmt.Errorf("VolumeBindingMode not set for StorageClass %q", className)
}
return *class.VolumeBindingMode == storage.VolumeBindingWaitForFirstConsumer, nil
}
// GetBindVolumeToClaim returns a new volume which is bound to given claim. In
// addition, it returns a bool which indicates whether we made modification on
// original volume.
func GetBindVolumeToClaim(volume *v1.PersistentVolume, claim *v1.PersistentVolumeClaim) (*v1.PersistentVolume, bool, error) {
dirty := false
// Check if the volume was already bound (either by user or by controller)
shouldSetBoundByController := false
if !IsVolumeBoundToClaim(volume, claim) {
shouldSetBoundByController = true
}
// The volume from method args can be pointing to watcher cache. We must not
// modify these, therefore create a copy.
volumeClone := volume.DeepCopy()
// Bind the volume to the claim if it is not bound yet
if volume.Spec.ClaimRef == nil ||
volume.Spec.ClaimRef.Name != claim.Name ||
volume.Spec.ClaimRef.Namespace != claim.Namespace ||
volume.Spec.ClaimRef.UID != claim.UID {
claimRef, err := reference.GetReference(scheme.Scheme, claim)
if err != nil {
return nil, false, fmt.Errorf("Unexpected error getting claim reference: %v", err)
}
volumeClone.Spec.ClaimRef = claimRef
dirty = true
}
// Set AnnBoundByController if it is not set yet
if shouldSetBoundByController && !metav1.HasAnnotation(volumeClone.ObjectMeta, AnnBoundByController) {
metav1.SetMetaDataAnnotation(&volumeClone.ObjectMeta, AnnBoundByController, "yes")
dirty = true
}
return volumeClone, dirty, nil
}
// IsVolumeBoundToClaim returns true, if given volume is pre-bound or bound
// to specific claim. Both claim.Name and claim.Namespace must be equal.
// If claim.UID is present in volume.Spec.ClaimRef, it must be equal too.
func IsVolumeBoundToClaim(volume *v1.PersistentVolume, claim *v1.PersistentVolumeClaim) bool {
if volume.Spec.ClaimRef == nil {
return false
}
if claim.Name != volume.Spec.ClaimRef.Name || claim.Namespace != volume.Spec.ClaimRef.Namespace {
return false
}
if volume.Spec.ClaimRef.UID != "" && claim.UID != volume.Spec.ClaimRef.UID {
return false
}
return true
}
// FindMatchingVolume goes through the list of volumes to find the best matching volume
// for the claim.
//
// This function is used by both the PV controller and scheduler.
//
// delayBinding is true only in the PV controller path. When set, prebound PVs are still returned
// as a match for the claim, but unbound PVs are skipped.
//
// node is set only in the scheduler path. When set, the PV node affinity is checked against
// the node's labels.
//
// excludedVolumes is only used in the scheduler path, and is needed for evaluating multiple
// unbound PVCs for a single Pod at one time. As each PVC finds a matching PV, the chosen
// PV needs to be excluded from future matching.
func FindMatchingVolume(
claim *v1.PersistentVolumeClaim,
volumes []*v1.PersistentVolume,
node *v1.Node,
excludedVolumes map[string]*v1.PersistentVolume,
delayBinding bool) (*v1.PersistentVolume, error) {
var smallestVolume *v1.PersistentVolume
var smallestVolumeQty resource.Quantity
requestedQty := claim.Spec.Resources.Requests[v1.ResourceName(v1.ResourceStorage)]
requestedClass := v1helper.GetPersistentVolumeClaimClass(claim)
var selector labels.Selector
if claim.Spec.Selector != nil {
internalSelector, err := metav1.LabelSelectorAsSelector(claim.Spec.Selector)
if err != nil {
// should be unreachable code due to validation
return nil, fmt.Errorf("error creating internal label selector for claim: %v: %v", claimToClaimKey(claim), err)
}
selector = internalSelector
}
// Go through all available volumes with two goals:
// - find a volume that is either pre-bound by user or dynamically
// provisioned for this claim. Because of this we need to loop through
// all volumes.
// - find the smallest matching one if there is no volume pre-bound to
// the claim.
for _, volume := range volumes {
if _, ok := excludedVolumes[volume.Name]; ok {
// Skip volumes in the excluded list
continue
}
if volume.Spec.ClaimRef != nil && !IsVolumeBoundToClaim(volume, claim) {
continue
}
volumeQty := volume.Spec.Capacity[v1.ResourceStorage]
if volumeQty.Cmp(requestedQty) < 0 {
continue
}
// filter out mismatching volumeModes
if CheckVolumeModeMismatches(&claim.Spec, &volume.Spec) {
continue
}
// check if PV's DeletionTimeStamp is set, if so, skip this volume.
if utilfeature.DefaultFeatureGate.Enabled(features.StorageObjectInUseProtection) {
if volume.ObjectMeta.DeletionTimestamp != nil {
continue
}
}
nodeAffinityValid := true
if node != nil {
// Scheduler path, check that the PV NodeAffinity
// is satisfied by the node
// volumeutil.CheckNodeAffinity is the most expensive call in this loop.
// We should check cheaper conditions first or consider optimizing this function.
err := volumeutil.CheckNodeAffinity(volume, node.Labels)
if err != nil {
nodeAffinityValid = false
}
}
if IsVolumeBoundToClaim(volume, claim) {
// If PV node affinity is invalid, return no match.
// This means the prebound PV (and therefore PVC)
// is not suitable for this node.
if !nodeAffinityValid {
return nil, nil
}
return volume, nil
}
if node == nil && delayBinding {
// PV controller does not bind this claim.
// Scheduler will handle binding unbound volumes
// Scheduler path will have node != nil
continue
}
// filter out:
// - volumes in non-available phase
// - volumes whose labels don't match the claim's selector, if specified
// - volumes in Class that is not requested
// - volumes whose NodeAffinity does not match the node
if volume.Status.Phase != v1.VolumeAvailable {
// We ignore volumes in non-available phase, because volumes that
// satisfies matching criteria will be updated to available, binding
// them now has high chance of encountering unnecessary failures
// due to API conflicts.
continue
} else if selector != nil && !selector.Matches(labels.Set(volume.Labels)) {
continue
}
if v1helper.GetPersistentVolumeClass(volume) != requestedClass {
continue
}
if !nodeAffinityValid {
continue
}
if node != nil {
// Scheduler path
// Check that the access modes match
if !CheckAccessModes(claim, volume) {
continue
}
}
if smallestVolume == nil || smallestVolumeQty.Cmp(volumeQty) > 0 {
smallestVolume = volume
smallestVolumeQty = volumeQty
}
}
if smallestVolume != nil {
// Found a matching volume
return smallestVolume, nil
}
return nil, nil
}
// CheckVolumeModeMismatches is a convenience method that checks volumeMode for PersistentVolume
// and PersistentVolumeClaims
func CheckVolumeModeMismatches(pvcSpec *v1.PersistentVolumeClaimSpec, pvSpec *v1.PersistentVolumeSpec) bool {
// In HA upgrades, we cannot guarantee that the apiserver is on a version >= controller-manager.
// So we default a nil volumeMode to filesystem
requestedVolumeMode := v1.PersistentVolumeFilesystem
if pvcSpec.VolumeMode != nil {
requestedVolumeMode = *pvcSpec.VolumeMode
}
pvVolumeMode := v1.PersistentVolumeFilesystem
if pvSpec.VolumeMode != nil {
pvVolumeMode = *pvSpec.VolumeMode
}
return requestedVolumeMode != pvVolumeMode
}
// CheckAccessModes returns true if PV satisfies all the PVC's requested AccessModes
func CheckAccessModes(claim *v1.PersistentVolumeClaim, volume *v1.PersistentVolume) bool {
pvModesMap := map[v1.PersistentVolumeAccessMode]bool{}
for _, mode := range volume.Spec.AccessModes {
pvModesMap[mode] = true
}
for _, mode := range claim.Spec.AccessModes {
_, ok := pvModesMap[mode]
if !ok {
return false
}
}
return true
}
func claimToClaimKey(claim *v1.PersistentVolumeClaim) string {
return fmt.Sprintf("%s/%s", claim.Namespace, claim.Name)
}
// GetVolumeNodeAffinity returns a VolumeNodeAffinity for given key and value.
func GetVolumeNodeAffinity(key string, value string) *v1.VolumeNodeAffinity {
return &v1.VolumeNodeAffinity{
Required: &v1.NodeSelector{
NodeSelectorTerms: []v1.NodeSelectorTerm{
{
MatchExpressions: []v1.NodeSelectorRequirement{
{
Key: key,
Operator: v1.NodeSelectorOpIn,
Values: []string{value},
},
},
},
},
},
}
}

View File

@ -1,87 +0,0 @@
load("@io_bazel_rules_go//go:def.bzl", "go_library", "go_test")
go_library(
name = "go_default_library",
srcs = [
"scheduler_assume_cache.go",
"scheduler_binder.go",
"scheduler_binder_cache.go",
"scheduler_binder_fake.go",
],
importpath = "k8s.io/kubernetes/pkg/controller/volume/scheduling",
visibility = ["//visibility:public"],
deps = [
"//pkg/apis/core/v1/helper:go_default_library",
"//pkg/controller/volume/persistentvolume/util:go_default_library",
"//pkg/controller/volume/scheduling/metrics:go_default_library",
"//pkg/features:go_default_library",
"//pkg/volume/util:go_default_library",
"//staging/src/k8s.io/api/core/v1:go_default_library",
"//staging/src/k8s.io/api/storage/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/api/meta:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/apis/meta/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/labels:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/sets:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/wait:go_default_library",
"//staging/src/k8s.io/apiserver/pkg/storage/etcd3:go_default_library",
"//staging/src/k8s.io/apiserver/pkg/util/feature:go_default_library",
"//staging/src/k8s.io/client-go/informers/core/v1:go_default_library",
"//staging/src/k8s.io/client-go/informers/storage/v1:go_default_library",
"//staging/src/k8s.io/client-go/kubernetes:go_default_library",
"//staging/src/k8s.io/client-go/listers/storage/v1:go_default_library",
"//staging/src/k8s.io/client-go/tools/cache:go_default_library",
"//staging/src/k8s.io/csi-translation-lib:go_default_library",
"//staging/src/k8s.io/csi-translation-lib/plugins:go_default_library",
"//vendor/k8s.io/klog:go_default_library",
],
)
go_test(
name = "go_default_test",
srcs = [
"scheduler_assume_cache_test.go",
"scheduler_binder_cache_test.go",
"scheduler_binder_test.go",
],
embed = [":go_default_library"],
deps = [
"//pkg/controller:go_default_library",
"//pkg/controller/volume/persistentvolume/testing:go_default_library",
"//pkg/controller/volume/persistentvolume/util:go_default_library",
"//pkg/features:go_default_library",
"//staging/src/k8s.io/api/core/v1:go_default_library",
"//staging/src/k8s.io/api/storage/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/api/resource:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/apis/meta/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/types:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/diff:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/wait:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/watch:go_default_library",
"//staging/src/k8s.io/apiserver/pkg/util/feature:go_default_library",
"//staging/src/k8s.io/client-go/informers:go_default_library",
"//staging/src/k8s.io/client-go/informers/core/v1:go_default_library",
"//staging/src/k8s.io/client-go/informers/storage/v1:go_default_library",
"//staging/src/k8s.io/client-go/kubernetes:go_default_library",
"//staging/src/k8s.io/client-go/kubernetes/fake:go_default_library",
"//staging/src/k8s.io/client-go/testing:go_default_library",
"//staging/src/k8s.io/component-base/featuregate/testing:go_default_library",
"//vendor/k8s.io/klog:go_default_library",
],
)
filegroup(
name = "package-srcs",
srcs = glob(["**"]),
tags = ["automanaged"],
visibility = ["//visibility:private"],
)
filegroup(
name = "all-srcs",
srcs = [
":package-srcs",
"//pkg/controller/volume/scheduling/metrics:all-srcs",
],
tags = ["automanaged"],
visibility = ["//visibility:public"],
)

View File

@ -1,9 +0,0 @@
# See the OWNERS docs at https://go.k8s.io/owners
approvers:
- msau42
- cofyc
reviewers:
- msau42
- cofyc
- lichuqiang

View File

@ -1,26 +0,0 @@
load("@io_bazel_rules_go//go:def.bzl", "go_library")
go_library(
name = "go_default_library",
srcs = ["metrics.go"],
importpath = "k8s.io/kubernetes/pkg/controller/volume/scheduling/metrics",
visibility = ["//visibility:public"],
deps = [
"//staging/src/k8s.io/component-base/metrics:go_default_library",
"//staging/src/k8s.io/component-base/metrics/legacyregistry:go_default_library",
],
)
filegroup(
name = "package-srcs",
srcs = glob(["**"]),
tags = ["automanaged"],
visibility = ["//visibility:private"],
)
filegroup(
name = "all-srcs",
srcs = [":package-srcs"],
tags = ["automanaged"],
visibility = ["//visibility:public"],
)

View File

@ -1,67 +0,0 @@
/*
Copyright 2018 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 metrics
import (
"k8s.io/component-base/metrics"
"k8s.io/component-base/metrics/legacyregistry"
)
// VolumeSchedulerSubsystem - subsystem name used by scheduler
const VolumeSchedulerSubsystem = "scheduler_volume"
var (
// VolumeBindingRequestSchedulerBinderCache tracks the number of volume binder cache operations.
VolumeBindingRequestSchedulerBinderCache = metrics.NewCounterVec(
&metrics.CounterOpts{
Subsystem: VolumeSchedulerSubsystem,
Name: "binder_cache_requests_total",
Help: "Total number for request volume binding cache",
StabilityLevel: metrics.ALPHA,
},
[]string{"operation"},
)
// VolumeSchedulingStageLatency tracks the latency of volume scheduling operations.
VolumeSchedulingStageLatency = metrics.NewHistogramVec(
&metrics.HistogramOpts{
Subsystem: VolumeSchedulerSubsystem,
Name: "scheduling_duration_seconds",
Help: "Volume scheduling stage latency",
Buckets: metrics.ExponentialBuckets(1000, 2, 15),
StabilityLevel: metrics.ALPHA,
},
[]string{"operation"},
)
// VolumeSchedulingStageFailed tracks the number of failed volume scheduling operations.
VolumeSchedulingStageFailed = metrics.NewCounterVec(
&metrics.CounterOpts{
Subsystem: VolumeSchedulerSubsystem,
Name: "scheduling_stage_error_total",
Help: "Volume scheduling stage error count",
StabilityLevel: metrics.ALPHA,
},
[]string{"operation"},
)
)
// RegisterVolumeSchedulingMetrics is used for scheduler, because the volume binding cache is a library
// used by scheduler process.
func RegisterVolumeSchedulingMetrics() {
legacyregistry.MustRegister(VolumeBindingRequestSchedulerBinderCache)
legacyregistry.MustRegister(VolumeSchedulingStageLatency)
legacyregistry.MustRegister(VolumeSchedulingStageFailed)
}

View File

@ -1,451 +0,0 @@
/*
Copyright 2017 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 scheduling
import (
"fmt"
"strconv"
"sync"
"k8s.io/klog"
"k8s.io/api/core/v1"
"k8s.io/apimachinery/pkg/api/meta"
"k8s.io/client-go/tools/cache"
)
// AssumeCache is a cache on top of the informer that allows for updating
// objects outside of informer events and also restoring the informer
// cache's version of the object. Objects are assumed to be
// Kubernetes API objects that implement meta.Interface
type AssumeCache interface {
// Assume updates the object in-memory only
Assume(obj interface{}) error
// Restore the informer cache's version of the object
Restore(objName string)
// Get the object by name
Get(objName string) (interface{}, error)
// Get the API object by name
GetAPIObj(objName string) (interface{}, error)
// List all the objects in the cache
List(indexObj interface{}) []interface{}
}
type errWrongType struct {
typeName string
object interface{}
}
func (e *errWrongType) Error() string {
return fmt.Sprintf("could not convert object to type %v: %+v", e.typeName, e.object)
}
type errNotFound struct {
typeName string
objectName string
}
func (e *errNotFound) Error() string {
return fmt.Sprintf("could not find %v %q", e.typeName, e.objectName)
}
type errObjectName struct {
detailedErr error
}
func (e *errObjectName) Error() string {
return fmt.Sprintf("failed to get object name: %v", e.detailedErr)
}
// assumeCache stores two pointers to represent a single object:
// * The pointer to the informer object.
// * The pointer to the latest object, which could be the same as
// the informer object, or an in-memory object.
//
// An informer update always overrides the latest object pointer.
//
// Assume() only updates the latest object pointer.
// Restore() sets the latest object pointer back to the informer object.
// Get/List() always returns the latest object pointer.
type assumeCache struct {
// Synchronizes updates to store
rwMutex sync.RWMutex
// describes the object stored
description string
// Stores objInfo pointers
store cache.Indexer
// Index function for object
indexFunc cache.IndexFunc
indexName string
}
type objInfo struct {
// name of the object
name string
// Latest version of object could be cached-only or from informer
latestObj interface{}
// Latest object from informer
apiObj interface{}
}
func objInfoKeyFunc(obj interface{}) (string, error) {
objInfo, ok := obj.(*objInfo)
if !ok {
return "", &errWrongType{"objInfo", obj}
}
return objInfo.name, nil
}
func (c *assumeCache) objInfoIndexFunc(obj interface{}) ([]string, error) {
objInfo, ok := obj.(*objInfo)
if !ok {
return []string{""}, &errWrongType{"objInfo", obj}
}
return c.indexFunc(objInfo.latestObj)
}
// NewAssumeCache creates an assume cache for general objects.
func NewAssumeCache(informer cache.SharedIndexInformer, description, indexName string, indexFunc cache.IndexFunc) AssumeCache {
c := &assumeCache{
description: description,
indexFunc: indexFunc,
indexName: indexName,
}
c.store = cache.NewIndexer(objInfoKeyFunc, cache.Indexers{indexName: c.objInfoIndexFunc})
// Unit tests don't use informers
if informer != nil {
informer.AddEventHandler(
cache.ResourceEventHandlerFuncs{
AddFunc: c.add,
UpdateFunc: c.update,
DeleteFunc: c.delete,
},
)
}
return c
}
func (c *assumeCache) add(obj interface{}) {
if obj == nil {
return
}
name, err := cache.MetaNamespaceKeyFunc(obj)
if err != nil {
klog.Errorf("add failed: %v", &errObjectName{err})
return
}
c.rwMutex.Lock()
defer c.rwMutex.Unlock()
if objInfo, _ := c.getObjInfo(name); objInfo != nil {
newVersion, err := c.getObjVersion(name, obj)
if err != nil {
klog.Errorf("add: couldn't get object version: %v", err)
return
}
storedVersion, err := c.getObjVersion(name, objInfo.latestObj)
if err != nil {
klog.Errorf("add: couldn't get stored object version: %v", err)
return
}
// Only update object if version is newer.
// This is so we don't override assumed objects due to informer resync.
if newVersion <= storedVersion {
klog.V(10).Infof("Skip adding %v %v to assume cache because version %v is not newer than %v", c.description, name, newVersion, storedVersion)
return
}
}
objInfo := &objInfo{name: name, latestObj: obj, apiObj: obj}
if err = c.store.Update(objInfo); err != nil {
klog.Warningf("got error when updating stored object : %v", err)
} else {
klog.V(10).Infof("Adding %v %v to assume cache: %+v ", c.description, name, obj)
}
}
func (c *assumeCache) update(oldObj interface{}, newObj interface{}) {
c.add(newObj)
}
func (c *assumeCache) delete(obj interface{}) {
if obj == nil {
return
}
name, err := cache.MetaNamespaceKeyFunc(obj)
if err != nil {
klog.Errorf("delete failed: %v", &errObjectName{err})
return
}
c.rwMutex.Lock()
defer c.rwMutex.Unlock()
objInfo := &objInfo{name: name}
err = c.store.Delete(objInfo)
if err != nil {
klog.Errorf("delete: failed to delete %v %v: %v", c.description, name, err)
}
}
func (c *assumeCache) getObjVersion(name string, obj interface{}) (int64, error) {
objAccessor, err := meta.Accessor(obj)
if err != nil {
return -1, err
}
objResourceVersion, err := strconv.ParseInt(objAccessor.GetResourceVersion(), 10, 64)
if err != nil {
return -1, fmt.Errorf("error parsing ResourceVersion %q for %v %q: %s", objAccessor.GetResourceVersion(), c.description, name, err)
}
return objResourceVersion, nil
}
func (c *assumeCache) getObjInfo(name string) (*objInfo, error) {
obj, ok, err := c.store.GetByKey(name)
if err != nil {
return nil, err
}
if !ok {
return nil, &errNotFound{c.description, name}
}
objInfo, ok := obj.(*objInfo)
if !ok {
return nil, &errWrongType{"objInfo", obj}
}
return objInfo, nil
}
func (c *assumeCache) Get(objName string) (interface{}, error) {
c.rwMutex.RLock()
defer c.rwMutex.RUnlock()
objInfo, err := c.getObjInfo(objName)
if err != nil {
return nil, err
}
return objInfo.latestObj, nil
}
func (c *assumeCache) GetAPIObj(objName string) (interface{}, error) {
c.rwMutex.RLock()
defer c.rwMutex.RUnlock()
objInfo, err := c.getObjInfo(objName)
if err != nil {
return nil, err
}
return objInfo.apiObj, nil
}
func (c *assumeCache) List(indexObj interface{}) []interface{} {
c.rwMutex.RLock()
defer c.rwMutex.RUnlock()
allObjs := []interface{}{}
objs, err := c.store.Index(c.indexName, &objInfo{latestObj: indexObj})
if err != nil {
klog.Errorf("list index error: %v", err)
return nil
}
for _, obj := range objs {
objInfo, ok := obj.(*objInfo)
if !ok {
klog.Errorf("list error: %v", &errWrongType{"objInfo", obj})
continue
}
allObjs = append(allObjs, objInfo.latestObj)
}
return allObjs
}
func (c *assumeCache) Assume(obj interface{}) error {
name, err := cache.MetaNamespaceKeyFunc(obj)
if err != nil {
return &errObjectName{err}
}
c.rwMutex.Lock()
defer c.rwMutex.Unlock()
objInfo, err := c.getObjInfo(name)
if err != nil {
return err
}
newVersion, err := c.getObjVersion(name, obj)
if err != nil {
return err
}
storedVersion, err := c.getObjVersion(name, objInfo.latestObj)
if err != nil {
return err
}
if newVersion < storedVersion {
return fmt.Errorf("%v %q is out of sync (stored: %d, assume: %d)", c.description, name, storedVersion, newVersion)
}
// Only update the cached object
objInfo.latestObj = obj
klog.V(4).Infof("Assumed %v %q, version %v", c.description, name, newVersion)
return nil
}
func (c *assumeCache) Restore(objName string) {
c.rwMutex.Lock()
defer c.rwMutex.Unlock()
objInfo, err := c.getObjInfo(objName)
if err != nil {
// This could be expected if object got deleted
klog.V(5).Infof("Restore %v %q warning: %v", c.description, objName, err)
} else {
objInfo.latestObj = objInfo.apiObj
klog.V(4).Infof("Restored %v %q", c.description, objName)
}
}
// PVAssumeCache is a AssumeCache for PersistentVolume objects
type PVAssumeCache interface {
AssumeCache
GetPV(pvName string) (*v1.PersistentVolume, error)
GetAPIPV(pvName string) (*v1.PersistentVolume, error)
ListPVs(storageClassName string) []*v1.PersistentVolume
}
type pvAssumeCache struct {
AssumeCache
}
func pvStorageClassIndexFunc(obj interface{}) ([]string, error) {
if pv, ok := obj.(*v1.PersistentVolume); ok {
return []string{pv.Spec.StorageClassName}, nil
}
return []string{""}, fmt.Errorf("object is not a v1.PersistentVolume: %v", obj)
}
// NewPVAssumeCache creates a PV assume cache.
func NewPVAssumeCache(informer cache.SharedIndexInformer) PVAssumeCache {
return &pvAssumeCache{NewAssumeCache(informer, "v1.PersistentVolume", "storageclass", pvStorageClassIndexFunc)}
}
func (c *pvAssumeCache) GetPV(pvName string) (*v1.PersistentVolume, error) {
obj, err := c.Get(pvName)
if err != nil {
return nil, err
}
pv, ok := obj.(*v1.PersistentVolume)
if !ok {
return nil, &errWrongType{"v1.PersistentVolume", obj}
}
return pv, nil
}
func (c *pvAssumeCache) GetAPIPV(pvName string) (*v1.PersistentVolume, error) {
obj, err := c.GetAPIObj(pvName)
if err != nil {
return nil, err
}
pv, ok := obj.(*v1.PersistentVolume)
if !ok {
return nil, &errWrongType{"v1.PersistentVolume", obj}
}
return pv, nil
}
func (c *pvAssumeCache) ListPVs(storageClassName string) []*v1.PersistentVolume {
objs := c.List(&v1.PersistentVolume{
Spec: v1.PersistentVolumeSpec{
StorageClassName: storageClassName,
},
})
pvs := []*v1.PersistentVolume{}
for _, obj := range objs {
pv, ok := obj.(*v1.PersistentVolume)
if !ok {
klog.Errorf("ListPVs: %v", &errWrongType{"v1.PersistentVolume", obj})
continue
}
pvs = append(pvs, pv)
}
return pvs
}
// PVCAssumeCache is a AssumeCache for PersistentVolumeClaim objects
type PVCAssumeCache interface {
AssumeCache
// GetPVC returns the PVC from the cache with given pvcKey.
// pvcKey is the result of MetaNamespaceKeyFunc on PVC obj
GetPVC(pvcKey string) (*v1.PersistentVolumeClaim, error)
GetAPIPVC(pvcKey string) (*v1.PersistentVolumeClaim, error)
}
type pvcAssumeCache struct {
AssumeCache
}
// NewPVCAssumeCache creates a PVC assume cache.
func NewPVCAssumeCache(informer cache.SharedIndexInformer) PVCAssumeCache {
return &pvcAssumeCache{NewAssumeCache(informer, "v1.PersistentVolumeClaim", "namespace", cache.MetaNamespaceIndexFunc)}
}
func (c *pvcAssumeCache) GetPVC(pvcKey string) (*v1.PersistentVolumeClaim, error) {
obj, err := c.Get(pvcKey)
if err != nil {
return nil, err
}
pvc, ok := obj.(*v1.PersistentVolumeClaim)
if !ok {
return nil, &errWrongType{"v1.PersistentVolumeClaim", obj}
}
return pvc, nil
}
func (c *pvcAssumeCache) GetAPIPVC(pvcKey string) (*v1.PersistentVolumeClaim, error) {
obj, err := c.GetAPIObj(pvcKey)
if err != nil {
return nil, err
}
pvc, ok := obj.(*v1.PersistentVolumeClaim)
if !ok {
return nil, &errWrongType{"v1.PersistentVolumeClaim", obj}
}
return pvc, nil
}

View File

@ -1,948 +0,0 @@
/*
Copyright 2017 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 scheduling
import (
"context"
"fmt"
"sort"
"strings"
"time"
v1 "k8s.io/api/core/v1"
storagev1 "k8s.io/api/storage/v1"
metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
"k8s.io/apimachinery/pkg/labels"
"k8s.io/apimachinery/pkg/util/sets"
"k8s.io/apimachinery/pkg/util/wait"
"k8s.io/apiserver/pkg/storage/etcd3"
utilfeature "k8s.io/apiserver/pkg/util/feature"
coreinformers "k8s.io/client-go/informers/core/v1"
storageinformers "k8s.io/client-go/informers/storage/v1"
clientset "k8s.io/client-go/kubernetes"
storagelisters "k8s.io/client-go/listers/storage/v1"
csitrans "k8s.io/csi-translation-lib"
csiplugins "k8s.io/csi-translation-lib/plugins"
"k8s.io/klog"
v1helper "k8s.io/kubernetes/pkg/apis/core/v1/helper"
pvutil "k8s.io/kubernetes/pkg/controller/volume/persistentvolume/util"
"k8s.io/kubernetes/pkg/controller/volume/scheduling/metrics"
"k8s.io/kubernetes/pkg/features"
volumeutil "k8s.io/kubernetes/pkg/volume/util"
)
// ConflictReason is used for the special strings which explain why
// volume binding is impossible for a node.
type ConflictReason string
// ConflictReasons contains all reasons that explain why volume binding is impossible for a node.
type ConflictReasons []ConflictReason
func (reasons ConflictReasons) Len() int { return len(reasons) }
func (reasons ConflictReasons) Less(i, j int) bool { return reasons[i] < reasons[j] }
func (reasons ConflictReasons) Swap(i, j int) { reasons[i], reasons[j] = reasons[j], reasons[i] }
const (
// ErrReasonBindConflict is used for VolumeBindingNoMatch predicate error.
ErrReasonBindConflict ConflictReason = "node(s) didn't find available persistent volumes to bind"
// ErrReasonNodeConflict is used for VolumeNodeAffinityConflict predicate error.
ErrReasonNodeConflict ConflictReason = "node(s) had volume node affinity conflict"
)
// InTreeToCSITranslator contains methods required to check migratable status
// and perform translations from InTree PV's to CSI
type InTreeToCSITranslator interface {
IsPVMigratable(pv *v1.PersistentVolume) bool
GetInTreePluginNameFromSpec(pv *v1.PersistentVolume, vol *v1.Volume) (string, error)
TranslateInTreePVToCSI(pv *v1.PersistentVolume) (*v1.PersistentVolume, error)
}
// SchedulerVolumeBinder is used by the scheduler to handle PVC/PV binding
// and dynamic provisioning. The binding decisions are integrated into the pod scheduling
// workflow so that the PV NodeAffinity is also considered along with the pod's other
// scheduling requirements.
//
// This integrates into the existing default scheduler workflow as follows:
// 1. The scheduler takes a Pod off the scheduler queue and processes it serially:
// a. Invokes all predicate functions, parallelized across nodes. FindPodVolumes() is invoked here.
// b. Invokes all priority functions. Future/TBD
// c. Selects the best node for the Pod.
// d. Cache the node selection for the Pod. AssumePodVolumes() is invoked here.
// i. If PVC binding is required, cache in-memory only:
// * For manual binding: update PV objects for prebinding to the corresponding PVCs.
// * For dynamic provisioning: update PVC object with a selected node from c)
// * For the pod, which PVCs and PVs need API updates.
// ii. Afterwards, the main scheduler caches the Pod->Node binding in the scheduler's pod cache,
// This is handled in the scheduler and not here.
// e. Asynchronously bind volumes and pod in a separate goroutine
// i. BindPodVolumes() is called first. It makes all the necessary API updates and waits for
// PV controller to fully bind and provision the PVCs. If binding fails, the Pod is sent
// back through the scheduler.
// ii. After BindPodVolumes() is complete, then the scheduler does the final Pod->Node binding.
// 2. Once all the assume operations are done in d), the scheduler processes the next Pod in the scheduler queue
// while the actual binding operation occurs in the background.
type SchedulerVolumeBinder interface {
// FindPodVolumes checks if all of a Pod's PVCs can be satisfied by the node.
//
// If a PVC is bound, it checks if the PV's NodeAffinity matches the Node.
// Otherwise, it tries to find an available PV to bind to the PVC.
//
// It returns an error when something went wrong or a list of reasons why the node is
// (currently) not usable for the pod.
//
// This function is called by the volume binding scheduler predicate and can be called in parallel
FindPodVolumes(pod *v1.Pod, node *v1.Node) (reasons ConflictReasons, err error)
// AssumePodVolumes will:
// 1. Take the PV matches for unbound PVCs and update the PV cache assuming
// that the PV is prebound to the PVC.
// 2. Take the PVCs that need provisioning and update the PVC cache with related
// annotations set.
//
// It returns true if all volumes are fully bound
//
// This function will modify assumedPod with the node name.
// This function is called serially.
AssumePodVolumes(assumedPod *v1.Pod, nodeName string) (allFullyBound bool, err error)
// BindPodVolumes will:
// 1. Initiate the volume binding by making the API call to prebind the PV
// to its matching PVC.
// 2. Trigger the volume provisioning by making the API call to set related
// annotations on the PVC
// 3. Wait for PVCs to be completely bound by the PV controller
//
// This function can be called in parallel.
BindPodVolumes(assumedPod *v1.Pod) error
// GetBindingsCache returns the cache used (if any) to store volume binding decisions.
GetBindingsCache() PodBindingCache
// DeletePodBindings will delete pod's bindingDecisions in podBindingCache.
DeletePodBindings(pod *v1.Pod)
}
type volumeBinder struct {
kubeClient clientset.Interface
classLister storagelisters.StorageClassLister
nodeInformer coreinformers.NodeInformer
csiNodeInformer storageinformers.CSINodeInformer
pvcCache PVCAssumeCache
pvCache PVAssumeCache
// Stores binding decisions that were made in FindPodVolumes for use in AssumePodVolumes.
// AssumePodVolumes modifies the bindings again for use in BindPodVolumes.
podBindingCache PodBindingCache
// Amount of time to wait for the bind operation to succeed
bindTimeout time.Duration
translator InTreeToCSITranslator
}
// NewVolumeBinder sets up all the caches needed for the scheduler to make volume binding decisions.
func NewVolumeBinder(
kubeClient clientset.Interface,
nodeInformer coreinformers.NodeInformer,
csiNodeInformer storageinformers.CSINodeInformer,
pvcInformer coreinformers.PersistentVolumeClaimInformer,
pvInformer coreinformers.PersistentVolumeInformer,
storageClassInformer storageinformers.StorageClassInformer,
bindTimeout time.Duration) SchedulerVolumeBinder {
b := &volumeBinder{
kubeClient: kubeClient,
classLister: storageClassInformer.Lister(),
nodeInformer: nodeInformer,
csiNodeInformer: csiNodeInformer,
pvcCache: NewPVCAssumeCache(pvcInformer.Informer()),
pvCache: NewPVAssumeCache(pvInformer.Informer()),
podBindingCache: NewPodBindingCache(),
bindTimeout: bindTimeout,
translator: csitrans.New(),
}
return b
}
func (b *volumeBinder) GetBindingsCache() PodBindingCache {
return b.podBindingCache
}
// DeletePodBindings will delete pod's bindingDecisions in podBindingCache.
func (b *volumeBinder) DeletePodBindings(pod *v1.Pod) {
cache := b.podBindingCache
if pod != nil {
cache.DeleteBindings(pod)
}
}
// FindPodVolumes caches the matching PVs and PVCs to provision per node in podBindingCache.
// This method intentionally takes in a *v1.Node object instead of using volumebinder.nodeInformer.
// That's necessary because some operations will need to pass in to the predicate fake node objects.
func (b *volumeBinder) FindPodVolumes(pod *v1.Pod, node *v1.Node) (reasons ConflictReasons, err error) {
podName := getPodName(pod)
// Warning: Below log needs high verbosity as it can be printed several times (#60933).
klog.V(5).Infof("FindPodVolumes for pod %q, node %q", podName, node.Name)
// Initialize to true for pods that don't have volumes. These
// booleans get translated into reason strings when the function
// returns without an error.
unboundVolumesSatisfied := true
boundVolumesSatisfied := true
defer func() {
if err != nil {
return
}
if !boundVolumesSatisfied {
reasons = append(reasons, ErrReasonNodeConflict)
}
if !unboundVolumesSatisfied {
reasons = append(reasons, ErrReasonBindConflict)
}
}()
start := time.Now()
defer func() {
metrics.VolumeSchedulingStageLatency.WithLabelValues("predicate").Observe(time.Since(start).Seconds())
if err != nil {
metrics.VolumeSchedulingStageFailed.WithLabelValues("predicate").Inc()
}
}()
var (
matchedBindings []*bindingInfo
provisionedClaims []*v1.PersistentVolumeClaim
)
defer func() {
// We recreate bindings for each new schedule loop.
if len(matchedBindings) == 0 && len(provisionedClaims) == 0 {
// Clear cache if no claims to bind or provision for this node.
b.podBindingCache.ClearBindings(pod, node.Name)
return
}
// Although we do not distinguish nil from empty in this function, for
// easier testing, we normalize empty to nil.
if len(matchedBindings) == 0 {
matchedBindings = nil
}
if len(provisionedClaims) == 0 {
provisionedClaims = nil
}
// Mark cache with all matched and provisioned claims for this node
b.podBindingCache.UpdateBindings(pod, node.Name, matchedBindings, provisionedClaims)
}()
// The pod's volumes need to be processed in one call to avoid the race condition where
// volumes can get bound/provisioned in between calls.
boundClaims, claimsToBind, unboundClaimsImmediate, err := b.getPodVolumes(pod)
if err != nil {
return nil, err
}
// Immediate claims should be bound
if len(unboundClaimsImmediate) > 0 {
return nil, fmt.Errorf("pod has unbound immediate PersistentVolumeClaims")
}
// Check PV node affinity on bound volumes
if len(boundClaims) > 0 {
boundVolumesSatisfied, err = b.checkBoundClaims(boundClaims, node, podName)
if err != nil {
return nil, err
}
}
// Find matching volumes and node for unbound claims
if len(claimsToBind) > 0 {
var (
claimsToFindMatching []*v1.PersistentVolumeClaim
claimsToProvision []*v1.PersistentVolumeClaim
)
// Filter out claims to provision
for _, claim := range claimsToBind {
if selectedNode, ok := claim.Annotations[pvutil.AnnSelectedNode]; ok {
if selectedNode != node.Name {
// Fast path, skip unmatched node.
unboundVolumesSatisfied = false
return
}
claimsToProvision = append(claimsToProvision, claim)
} else {
claimsToFindMatching = append(claimsToFindMatching, claim)
}
}
// Find matching volumes
if len(claimsToFindMatching) > 0 {
var unboundClaims []*v1.PersistentVolumeClaim
unboundVolumesSatisfied, matchedBindings, unboundClaims, err = b.findMatchingVolumes(pod, claimsToFindMatching, node)
if err != nil {
return nil, err
}
claimsToProvision = append(claimsToProvision, unboundClaims...)
}
// Check for claims to provision
if len(claimsToProvision) > 0 {
unboundVolumesSatisfied, provisionedClaims, err = b.checkVolumeProvisions(pod, claimsToProvision, node)
if err != nil {
return nil, err
}
}
}
return
}
// AssumePodVolumes will take the cached matching PVs and PVCs to provision
// in podBindingCache for the chosen node, and:
// 1. Update the pvCache with the new prebound PV.
// 2. Update the pvcCache with the new PVCs with annotations set
// 3. Update podBindingCache again with cached API updates for PVs and PVCs.
func (b *volumeBinder) AssumePodVolumes(assumedPod *v1.Pod, nodeName string) (allFullyBound bool, err error) {
podName := getPodName(assumedPod)
klog.V(4).Infof("AssumePodVolumes for pod %q, node %q", podName, nodeName)
start := time.Now()
defer func() {
metrics.VolumeSchedulingStageLatency.WithLabelValues("assume").Observe(time.Since(start).Seconds())
if err != nil {
metrics.VolumeSchedulingStageFailed.WithLabelValues("assume").Inc()
}
}()
if allBound := b.arePodVolumesBound(assumedPod); allBound {
klog.V(4).Infof("AssumePodVolumes for pod %q, node %q: all PVCs bound and nothing to do", podName, nodeName)
return true, nil
}
assumedPod.Spec.NodeName = nodeName
claimsToBind := b.podBindingCache.GetBindings(assumedPod, nodeName)
claimsToProvision := b.podBindingCache.GetProvisionedPVCs(assumedPod, nodeName)
// Assume PV
newBindings := []*bindingInfo{}
for _, binding := range claimsToBind {
newPV, dirty, err := pvutil.GetBindVolumeToClaim(binding.pv, binding.pvc)
klog.V(5).Infof("AssumePodVolumes: GetBindVolumeToClaim for pod %q, PV %q, PVC %q. newPV %p, dirty %v, err: %v",
podName,
binding.pv.Name,
binding.pvc.Name,
newPV,
dirty,
err)
if err != nil {
b.revertAssumedPVs(newBindings)
return false, err
}
// TODO: can we assume everytime?
if dirty {
err = b.pvCache.Assume(newPV)
if err != nil {
b.revertAssumedPVs(newBindings)
return false, err
}
}
newBindings = append(newBindings, &bindingInfo{pv: newPV, pvc: binding.pvc})
}
// Assume PVCs
newProvisionedPVCs := []*v1.PersistentVolumeClaim{}
for _, claim := range claimsToProvision {
// The claims from method args can be pointing to watcher cache. We must not
// modify these, therefore create a copy.
claimClone := claim.DeepCopy()
metav1.SetMetaDataAnnotation(&claimClone.ObjectMeta, pvutil.AnnSelectedNode, nodeName)
err = b.pvcCache.Assume(claimClone)
if err != nil {
b.revertAssumedPVs(newBindings)
b.revertAssumedPVCs(newProvisionedPVCs)
return
}
newProvisionedPVCs = append(newProvisionedPVCs, claimClone)
}
// Update cache with the assumed pvcs and pvs
// Even if length is zero, update the cache with an empty slice to indicate that no
// operations are needed
b.podBindingCache.UpdateBindings(assumedPod, nodeName, newBindings, newProvisionedPVCs)
return
}
// BindPodVolumes gets the cached bindings and PVCs to provision in podBindingCache,
// makes the API update for those PVs/PVCs, and waits for the PVCs to be completely bound
// by the PV controller.
func (b *volumeBinder) BindPodVolumes(assumedPod *v1.Pod) (err error) {
podName := getPodName(assumedPod)
klog.V(4).Infof("BindPodVolumes for pod %q, node %q", podName, assumedPod.Spec.NodeName)
start := time.Now()
defer func() {
metrics.VolumeSchedulingStageLatency.WithLabelValues("bind").Observe(time.Since(start).Seconds())
if err != nil {
metrics.VolumeSchedulingStageFailed.WithLabelValues("bind").Inc()
}
}()
bindings := b.podBindingCache.GetBindings(assumedPod, assumedPod.Spec.NodeName)
claimsToProvision := b.podBindingCache.GetProvisionedPVCs(assumedPod, assumedPod.Spec.NodeName)
// Start API operations
err = b.bindAPIUpdate(podName, bindings, claimsToProvision)
if err != nil {
return err
}
err = wait.Poll(time.Second, b.bindTimeout, func() (bool, error) {
b, err := b.checkBindings(assumedPod, bindings, claimsToProvision)
return b, err
})
if err != nil {
return fmt.Errorf("Failed to bind volumes: %v", err)
}
return nil
}
func getPodName(pod *v1.Pod) string {
return pod.Namespace + "/" + pod.Name
}
func getPVCName(pvc *v1.PersistentVolumeClaim) string {
return pvc.Namespace + "/" + pvc.Name
}
// bindAPIUpdate gets the cached bindings and PVCs to provision in podBindingCache
// and makes the API update for those PVs/PVCs.
func (b *volumeBinder) bindAPIUpdate(podName string, bindings []*bindingInfo, claimsToProvision []*v1.PersistentVolumeClaim) error {
if bindings == nil {
return fmt.Errorf("failed to get cached bindings for pod %q", podName)
}
if claimsToProvision == nil {
return fmt.Errorf("failed to get cached claims to provision for pod %q", podName)
}
lastProcessedBinding := 0
lastProcessedProvisioning := 0
defer func() {
// only revert assumed cached updates for volumes we haven't successfully bound
if lastProcessedBinding < len(bindings) {
b.revertAssumedPVs(bindings[lastProcessedBinding:])
}
// only revert assumed cached updates for claims we haven't updated,
if lastProcessedProvisioning < len(claimsToProvision) {
b.revertAssumedPVCs(claimsToProvision[lastProcessedProvisioning:])
}
}()
var (
binding *bindingInfo
i int
claim *v1.PersistentVolumeClaim
)
// Do the actual prebinding. Let the PV controller take care of the rest
// There is no API rollback if the actual binding fails
for _, binding = range bindings {
klog.V(5).Infof("bindAPIUpdate: Pod %q, binding PV %q to PVC %q", podName, binding.pv.Name, binding.pvc.Name)
// TODO: does it hurt if we make an api call and nothing needs to be updated?
claimKey := claimToClaimKey(binding.pvc)
klog.V(2).Infof("claim %q bound to volume %q", claimKey, binding.pv.Name)
newPV, err := b.kubeClient.CoreV1().PersistentVolumes().Update(context.TODO(), binding.pv, metav1.UpdateOptions{})
if err != nil {
klog.V(4).Infof("updating PersistentVolume[%s]: binding to %q failed: %v", binding.pv.Name, claimKey, err)
return err
}
klog.V(4).Infof("updating PersistentVolume[%s]: bound to %q", binding.pv.Name, claimKey)
// Save updated object from apiserver for later checking.
binding.pv = newPV
lastProcessedBinding++
}
// Update claims objects to trigger volume provisioning. Let the PV controller take care of the rest
// PV controller is expect to signal back by removing related annotations if actual provisioning fails
for i, claim = range claimsToProvision {
klog.V(5).Infof("bindAPIUpdate: Pod %q, PVC %q", podName, getPVCName(claim))
newClaim, err := b.kubeClient.CoreV1().PersistentVolumeClaims(claim.Namespace).Update(context.TODO(), claim, metav1.UpdateOptions{})
if err != nil {
return err
}
// Save updated object from apiserver for later checking.
claimsToProvision[i] = newClaim
lastProcessedProvisioning++
}
return nil
}
var (
versioner = etcd3.APIObjectVersioner{}
)
// checkBindings runs through all the PVCs in the Pod and checks:
// * if the PVC is fully bound
// * if there are any conditions that require binding to fail and be retried
//
// It returns true when all of the Pod's PVCs are fully bound, and error if
// binding (and scheduling) needs to be retried
// Note that it checks on API objects not PV/PVC cache, this is because
// PV/PVC cache can be assumed again in main scheduler loop, we must check
// latest state in API server which are shared with PV controller and
// provisioners
func (b *volumeBinder) checkBindings(pod *v1.Pod, bindings []*bindingInfo, claimsToProvision []*v1.PersistentVolumeClaim) (bool, error) {
podName := getPodName(pod)
if bindings == nil {
return false, fmt.Errorf("failed to get cached bindings for pod %q", podName)
}
if claimsToProvision == nil {
return false, fmt.Errorf("failed to get cached claims to provision for pod %q", podName)
}
node, err := b.nodeInformer.Lister().Get(pod.Spec.NodeName)
if err != nil {
return false, fmt.Errorf("failed to get node %q: %v", pod.Spec.NodeName, err)
}
csiNode, err := b.csiNodeInformer.Lister().Get(node.Name)
if err != nil {
// TODO: return the error once CSINode is created by default
klog.V(4).Infof("Could not get a CSINode object for the node %q: %v", node.Name, err)
}
// Check for any conditions that might require scheduling retry
// When pod is removed from scheduling queue because of deletion or any
// other reasons, binding operation should be cancelled. There is no need
// to check PV/PVC bindings any more.
// We check pod binding cache here which will be cleared when pod is
// removed from scheduling queue.
if b.podBindingCache.GetDecisions(pod) == nil {
return false, fmt.Errorf("pod %q does not exist any more", podName)
}
for _, binding := range bindings {
pv, err := b.pvCache.GetAPIPV(binding.pv.Name)
if err != nil {
return false, fmt.Errorf("failed to check binding: %v", err)
}
pvc, err := b.pvcCache.GetAPIPVC(getPVCName(binding.pvc))
if err != nil {
return false, fmt.Errorf("failed to check binding: %v", err)
}
// Because we updated PV in apiserver, skip if API object is older
// and wait for new API object propagated from apiserver.
if versioner.CompareResourceVersion(binding.pv, pv) > 0 {
return false, nil
}
pv, err = b.tryTranslatePVToCSI(pv, csiNode)
if err != nil {
return false, fmt.Errorf("failed to translate pv to csi: %v", err)
}
// Check PV's node affinity (the node might not have the proper label)
if err := volumeutil.CheckNodeAffinity(pv, node.Labels); err != nil {
return false, fmt.Errorf("pv %q node affinity doesn't match node %q: %v", pv.Name, node.Name, err)
}
// Check if pv.ClaimRef got dropped by unbindVolume()
if pv.Spec.ClaimRef == nil || pv.Spec.ClaimRef.UID == "" {
return false, fmt.Errorf("ClaimRef got reset for pv %q", pv.Name)
}
// Check if pvc is fully bound
if !b.isPVCFullyBound(pvc) {
return false, nil
}
}
for _, claim := range claimsToProvision {
pvc, err := b.pvcCache.GetAPIPVC(getPVCName(claim))
if err != nil {
return false, fmt.Errorf("failed to check provisioning pvc: %v", err)
}
// Because we updated PVC in apiserver, skip if API object is older
// and wait for new API object propagated from apiserver.
if versioner.CompareResourceVersion(claim, pvc) > 0 {
return false, nil
}
// Check if selectedNode annotation is still set
if pvc.Annotations == nil {
return false, fmt.Errorf("selectedNode annotation reset for PVC %q", pvc.Name)
}
selectedNode := pvc.Annotations[pvutil.AnnSelectedNode]
if selectedNode != pod.Spec.NodeName {
// If provisioner fails to provision a volume, selectedNode
// annotation will be removed to signal back to the scheduler to
// retry.
return false, fmt.Errorf("provisioning failed for PVC %q", pvc.Name)
}
// If the PVC is bound to a PV, check its node affinity
if pvc.Spec.VolumeName != "" {
pv, err := b.pvCache.GetAPIPV(pvc.Spec.VolumeName)
if err != nil {
if _, ok := err.(*errNotFound); ok {
// We tolerate NotFound error here, because PV is possibly
// not found because of API delay, we can check next time.
// And if PV does not exist because it's deleted, PVC will
// be unbound eventually.
return false, nil
}
return false, fmt.Errorf("failed to get pv %q from cache: %v", pvc.Spec.VolumeName, err)
}
pv, err = b.tryTranslatePVToCSI(pv, csiNode)
if err != nil {
return false, err
}
if err := volumeutil.CheckNodeAffinity(pv, node.Labels); err != nil {
return false, fmt.Errorf("pv %q node affinity doesn't match node %q: %v", pv.Name, node.Name, err)
}
}
// Check if pvc is fully bound
if !b.isPVCFullyBound(pvc) {
return false, nil
}
}
// All pvs and pvcs that we operated on are bound
klog.V(4).Infof("All PVCs for pod %q are bound", podName)
return true, nil
}
func (b *volumeBinder) isVolumeBound(namespace string, vol *v1.Volume) (bool, *v1.PersistentVolumeClaim, error) {
if vol.PersistentVolumeClaim == nil {
return true, nil, nil
}
pvcName := vol.PersistentVolumeClaim.ClaimName
return b.isPVCBound(namespace, pvcName)
}
func (b *volumeBinder) isPVCBound(namespace, pvcName string) (bool, *v1.PersistentVolumeClaim, error) {
claim := &v1.PersistentVolumeClaim{
ObjectMeta: metav1.ObjectMeta{
Name: pvcName,
Namespace: namespace,
},
}
pvcKey := getPVCName(claim)
pvc, err := b.pvcCache.GetPVC(pvcKey)
if err != nil || pvc == nil {
return false, nil, fmt.Errorf("error getting PVC %q: %v", pvcKey, err)
}
fullyBound := b.isPVCFullyBound(pvc)
if fullyBound {
klog.V(5).Infof("PVC %q is fully bound to PV %q", pvcKey, pvc.Spec.VolumeName)
} else {
if pvc.Spec.VolumeName != "" {
klog.V(5).Infof("PVC %q is not fully bound to PV %q", pvcKey, pvc.Spec.VolumeName)
} else {
klog.V(5).Infof("PVC %q is not bound", pvcKey)
}
}
return fullyBound, pvc, nil
}
func (b *volumeBinder) isPVCFullyBound(pvc *v1.PersistentVolumeClaim) bool {
return pvc.Spec.VolumeName != "" && metav1.HasAnnotation(pvc.ObjectMeta, pvutil.AnnBindCompleted)
}
// arePodVolumesBound returns true if all volumes are fully bound
func (b *volumeBinder) arePodVolumesBound(pod *v1.Pod) bool {
for _, vol := range pod.Spec.Volumes {
if isBound, _, _ := b.isVolumeBound(pod.Namespace, &vol); !isBound {
// Pod has at least one PVC that needs binding
return false
}
}
return true
}
// getPodVolumes returns a pod's PVCs separated into bound, unbound with delayed binding (including provisioning)
// and unbound with immediate binding (including prebound)
func (b *volumeBinder) getPodVolumes(pod *v1.Pod) (boundClaims []*v1.PersistentVolumeClaim, unboundClaimsDelayBinding []*v1.PersistentVolumeClaim, unboundClaimsImmediate []*v1.PersistentVolumeClaim, err error) {
boundClaims = []*v1.PersistentVolumeClaim{}
unboundClaimsImmediate = []*v1.PersistentVolumeClaim{}
unboundClaimsDelayBinding = []*v1.PersistentVolumeClaim{}
for _, vol := range pod.Spec.Volumes {
volumeBound, pvc, err := b.isVolumeBound(pod.Namespace, &vol)
if err != nil {
return nil, nil, nil, err
}
if pvc == nil {
continue
}
if volumeBound {
boundClaims = append(boundClaims, pvc)
} else {
delayBindingMode, err := pvutil.IsDelayBindingMode(pvc, b.classLister)
if err != nil {
return nil, nil, nil, err
}
// Prebound PVCs are treated as unbound immediate binding
if delayBindingMode && pvc.Spec.VolumeName == "" {
// Scheduler path
unboundClaimsDelayBinding = append(unboundClaimsDelayBinding, pvc)
} else {
// !delayBindingMode || pvc.Spec.VolumeName != ""
// Immediate binding should have already been bound
unboundClaimsImmediate = append(unboundClaimsImmediate, pvc)
}
}
}
return boundClaims, unboundClaimsDelayBinding, unboundClaimsImmediate, nil
}
func (b *volumeBinder) checkBoundClaims(claims []*v1.PersistentVolumeClaim, node *v1.Node, podName string) (bool, error) {
csiNode, err := b.csiNodeInformer.Lister().Get(node.Name)
if err != nil {
// TODO: return the error once CSINode is created by default
klog.V(4).Infof("Could not get a CSINode object for the node %q: %v", node.Name, err)
}
for _, pvc := range claims {
pvName := pvc.Spec.VolumeName
pv, err := b.pvCache.GetPV(pvName)
if err != nil {
return false, err
}
pv, err = b.tryTranslatePVToCSI(pv, csiNode)
if err != nil {
return false, err
}
err = volumeutil.CheckNodeAffinity(pv, node.Labels)
if err != nil {
klog.V(4).Infof("PersistentVolume %q, Node %q mismatch for Pod %q: %v", pvName, node.Name, podName, err)
return false, nil
}
klog.V(5).Infof("PersistentVolume %q, Node %q matches for Pod %q", pvName, node.Name, podName)
}
klog.V(4).Infof("All bound volumes for Pod %q match with Node %q", podName, node.Name)
return true, nil
}
// findMatchingVolumes tries to find matching volumes for given claims,
// and return unbound claims for further provision.
func (b *volumeBinder) findMatchingVolumes(pod *v1.Pod, claimsToBind []*v1.PersistentVolumeClaim, node *v1.Node) (foundMatches bool, bindings []*bindingInfo, unboundClaims []*v1.PersistentVolumeClaim, err error) {
podName := getPodName(pod)
// Sort all the claims by increasing size request to get the smallest fits
sort.Sort(byPVCSize(claimsToBind))
chosenPVs := map[string]*v1.PersistentVolume{}
foundMatches = true
for _, pvc := range claimsToBind {
// Get storage class name from each PVC
storageClassName := v1helper.GetPersistentVolumeClaimClass(pvc)
allPVs := b.pvCache.ListPVs(storageClassName)
pvcName := getPVCName(pvc)
// Find a matching PV
pv, err := pvutil.FindMatchingVolume(pvc, allPVs, node, chosenPVs, true)
if err != nil {
return false, nil, nil, err
}
if pv == nil {
klog.V(4).Infof("No matching volumes for Pod %q, PVC %q on node %q", podName, pvcName, node.Name)
unboundClaims = append(unboundClaims, pvc)
foundMatches = false
continue
}
// matching PV needs to be excluded so we don't select it again
chosenPVs[pv.Name] = pv
bindings = append(bindings, &bindingInfo{pv: pv, pvc: pvc})
klog.V(5).Infof("Found matching PV %q for PVC %q on node %q for pod %q", pv.Name, pvcName, node.Name, podName)
}
if foundMatches {
klog.V(4).Infof("Found matching volumes for pod %q on node %q", podName, node.Name)
}
return
}
// checkVolumeProvisions checks given unbound claims (the claims have gone through func
// findMatchingVolumes, and do not have matching volumes for binding), and return true
// if all of the claims are eligible for dynamic provision.
func (b *volumeBinder) checkVolumeProvisions(pod *v1.Pod, claimsToProvision []*v1.PersistentVolumeClaim, node *v1.Node) (provisionSatisfied bool, provisionedClaims []*v1.PersistentVolumeClaim, err error) {
podName := getPodName(pod)
provisionedClaims = []*v1.PersistentVolumeClaim{}
for _, claim := range claimsToProvision {
pvcName := getPVCName(claim)
className := v1helper.GetPersistentVolumeClaimClass(claim)
if className == "" {
return false, nil, fmt.Errorf("no class for claim %q", pvcName)
}
class, err := b.classLister.Get(className)
if err != nil {
return false, nil, fmt.Errorf("failed to find storage class %q", className)
}
provisioner := class.Provisioner
if provisioner == "" || provisioner == pvutil.NotSupportedProvisioner {
klog.V(4).Infof("storage class %q of claim %q does not support dynamic provisioning", className, pvcName)
return false, nil, nil
}
// Check if the node can satisfy the topology requirement in the class
if !v1helper.MatchTopologySelectorTerms(class.AllowedTopologies, labels.Set(node.Labels)) {
klog.V(4).Infof("Node %q cannot satisfy provisioning topology requirements of claim %q", node.Name, pvcName)
return false, nil, nil
}
// TODO: Check if capacity of the node domain in the storage class
// can satisfy resource requirement of given claim
provisionedClaims = append(provisionedClaims, claim)
}
klog.V(4).Infof("Provisioning for claims of pod %q that has no matching volumes on node %q ...", podName, node.Name)
return true, provisionedClaims, nil
}
func (b *volumeBinder) revertAssumedPVs(bindings []*bindingInfo) {
for _, bindingInfo := range bindings {
b.pvCache.Restore(bindingInfo.pv.Name)
}
}
func (b *volumeBinder) revertAssumedPVCs(claims []*v1.PersistentVolumeClaim) {
for _, claim := range claims {
b.pvcCache.Restore(getPVCName(claim))
}
}
type bindingInfo struct {
// Claim that needs to be bound
pvc *v1.PersistentVolumeClaim
// Proposed PV to bind to this claim
pv *v1.PersistentVolume
}
type byPVCSize []*v1.PersistentVolumeClaim
func (a byPVCSize) Len() int {
return len(a)
}
func (a byPVCSize) Swap(i, j int) {
a[i], a[j] = a[j], a[i]
}
func (a byPVCSize) Less(i, j int) bool {
iSize := a[i].Spec.Resources.Requests[v1.ResourceStorage]
jSize := a[j].Spec.Resources.Requests[v1.ResourceStorage]
// return true if iSize is less than jSize
return iSize.Cmp(jSize) == -1
}
func claimToClaimKey(claim *v1.PersistentVolumeClaim) string {
return fmt.Sprintf("%s/%s", claim.Namespace, claim.Name)
}
// isCSIMigrationOnForPlugin checks if CSI migrartion is enabled for a given plugin.
func isCSIMigrationOnForPlugin(pluginName string) bool {
switch pluginName {
case csiplugins.AWSEBSInTreePluginName:
return utilfeature.DefaultFeatureGate.Enabled(features.CSIMigrationAWS)
case csiplugins.GCEPDInTreePluginName:
return utilfeature.DefaultFeatureGate.Enabled(features.CSIMigrationGCE)
case csiplugins.AzureDiskInTreePluginName:
return utilfeature.DefaultFeatureGate.Enabled(features.CSIMigrationAzureDisk)
case csiplugins.CinderInTreePluginName:
return utilfeature.DefaultFeatureGate.Enabled(features.CSIMigrationOpenStack)
}
return false
}
// isPluginMigratedToCSIOnNode checks if an in-tree plugin has been migrated to a CSI driver on the node.
func isPluginMigratedToCSIOnNode(pluginName string, csiNode *storagev1.CSINode) bool {
if csiNode == nil {
return false
}
csiNodeAnn := csiNode.GetAnnotations()
if csiNodeAnn == nil {
return false
}
var mpaSet sets.String
mpa := csiNodeAnn[v1.MigratedPluginsAnnotationKey]
if len(mpa) == 0 {
mpaSet = sets.NewString()
} else {
tok := strings.Split(mpa, ",")
mpaSet = sets.NewString(tok...)
}
return mpaSet.Has(pluginName)
}
// tryTranslatePVToCSI will translate the in-tree PV to CSI if it meets the criteria. If not, it returns the unmodified in-tree PV.
func (b *volumeBinder) tryTranslatePVToCSI(pv *v1.PersistentVolume, csiNode *storagev1.CSINode) (*v1.PersistentVolume, error) {
if !b.translator.IsPVMigratable(pv) {
return pv, nil
}
if !utilfeature.DefaultFeatureGate.Enabled(features.CSIMigration) {
return pv, nil
}
pluginName, err := b.translator.GetInTreePluginNameFromSpec(pv, nil)
if err != nil {
return nil, fmt.Errorf("could not get plugin name from pv: %v", err)
}
if !isCSIMigrationOnForPlugin(pluginName) {
return pv, nil
}
if !isPluginMigratedToCSIOnNode(pluginName, csiNode) {
return pv, nil
}
transPV, err := b.translator.TranslateInTreePVToCSI(pv)
if err != nil {
return nil, fmt.Errorf("could not translate pv: %v", err)
}
return transPV, nil
}

View File

@ -1,167 +0,0 @@
/*
Copyright 2017 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 scheduling
import (
"sync"
v1 "k8s.io/api/core/v1"
"k8s.io/kubernetes/pkg/controller/volume/scheduling/metrics"
)
// PodBindingCache stores PV binding decisions per pod per node.
// Pod entries are removed when the Pod is deleted or updated to
// no longer be schedulable.
type PodBindingCache interface {
// UpdateBindings will update the cache with the given bindings for the
// pod and node.
UpdateBindings(pod *v1.Pod, node string, bindings []*bindingInfo, provisionings []*v1.PersistentVolumeClaim)
// ClearBindings will clear the cached bindings for the given pod and node.
ClearBindings(pod *v1.Pod, node string)
// GetBindings will return the cached bindings for the given pod and node.
// A nil return value means that the entry was not found. An empty slice
// means that no binding operations are needed.
GetBindings(pod *v1.Pod, node string) []*bindingInfo
// A nil return value means that the entry was not found. An empty slice
// means that no provisioning operations are needed.
GetProvisionedPVCs(pod *v1.Pod, node string) []*v1.PersistentVolumeClaim
// GetDecisions will return all cached decisions for the given pod.
GetDecisions(pod *v1.Pod) nodeDecisions
// DeleteBindings will remove all cached bindings and provisionings for the given pod.
// TODO: separate the func if it is needed to delete bindings/provisionings individually
DeleteBindings(pod *v1.Pod)
}
type podBindingCache struct {
// synchronizes bindingDecisions
rwMutex sync.RWMutex
// Key = pod name
// Value = nodeDecisions
bindingDecisions map[string]nodeDecisions
}
// Key = nodeName
// Value = bindings & provisioned PVCs of the node
type nodeDecisions map[string]nodeDecision
// A decision includes bindingInfo and provisioned PVCs of the node
type nodeDecision struct {
bindings []*bindingInfo
provisionings []*v1.PersistentVolumeClaim
}
// NewPodBindingCache creates a pod binding cache.
func NewPodBindingCache() PodBindingCache {
return &podBindingCache{bindingDecisions: map[string]nodeDecisions{}}
}
func (c *podBindingCache) GetDecisions(pod *v1.Pod) nodeDecisions {
c.rwMutex.RLock()
defer c.rwMutex.RUnlock()
podName := getPodName(pod)
decisions, ok := c.bindingDecisions[podName]
if !ok {
return nil
}
return decisions
}
func (c *podBindingCache) DeleteBindings(pod *v1.Pod) {
c.rwMutex.Lock()
defer c.rwMutex.Unlock()
podName := getPodName(pod)
if _, ok := c.bindingDecisions[podName]; ok {
delete(c.bindingDecisions, podName)
metrics.VolumeBindingRequestSchedulerBinderCache.WithLabelValues("delete").Inc()
}
}
func (c *podBindingCache) UpdateBindings(pod *v1.Pod, node string, bindings []*bindingInfo, pvcs []*v1.PersistentVolumeClaim) {
c.rwMutex.Lock()
defer c.rwMutex.Unlock()
podName := getPodName(pod)
decisions, ok := c.bindingDecisions[podName]
if !ok {
decisions = nodeDecisions{}
c.bindingDecisions[podName] = decisions
}
decision, ok := decisions[node]
if !ok {
decision = nodeDecision{
bindings: bindings,
provisionings: pvcs,
}
metrics.VolumeBindingRequestSchedulerBinderCache.WithLabelValues("add").Inc()
} else {
decision.bindings = bindings
decision.provisionings = pvcs
}
decisions[node] = decision
}
func (c *podBindingCache) GetBindings(pod *v1.Pod, node string) []*bindingInfo {
c.rwMutex.RLock()
defer c.rwMutex.RUnlock()
podName := getPodName(pod)
decisions, ok := c.bindingDecisions[podName]
if !ok {
return nil
}
decision, ok := decisions[node]
if !ok {
return nil
}
return decision.bindings
}
func (c *podBindingCache) GetProvisionedPVCs(pod *v1.Pod, node string) []*v1.PersistentVolumeClaim {
c.rwMutex.RLock()
defer c.rwMutex.RUnlock()
podName := getPodName(pod)
decisions, ok := c.bindingDecisions[podName]
if !ok {
return nil
}
decision, ok := decisions[node]
if !ok {
return nil
}
return decision.provisionings
}
func (c *podBindingCache) ClearBindings(pod *v1.Pod, node string) {
c.rwMutex.Lock()
defer c.rwMutex.Unlock()
podName := getPodName(pod)
decisions, ok := c.bindingDecisions[podName]
if !ok {
return
}
delete(decisions, node)
}

View File

@ -1,68 +0,0 @@
/*
Copyright 2017 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 scheduling
import "k8s.io/api/core/v1"
// FakeVolumeBinderConfig holds configurations for fake volume binder.
type FakeVolumeBinderConfig struct {
AllBound bool
FindReasons ConflictReasons
FindErr error
AssumeErr error
BindErr error
}
// NewFakeVolumeBinder sets up all the caches needed for the scheduler to make
// topology-aware volume binding decisions.
func NewFakeVolumeBinder(config *FakeVolumeBinderConfig) *FakeVolumeBinder {
return &FakeVolumeBinder{
config: config,
}
}
// FakeVolumeBinder represents a fake volume binder for testing.
type FakeVolumeBinder struct {
config *FakeVolumeBinderConfig
AssumeCalled bool
BindCalled bool
}
// FindPodVolumes implements SchedulerVolumeBinder.FindPodVolumes.
func (b *FakeVolumeBinder) FindPodVolumes(pod *v1.Pod, node *v1.Node) (reasons ConflictReasons, err error) {
return b.config.FindReasons, b.config.FindErr
}
// AssumePodVolumes implements SchedulerVolumeBinder.AssumePodVolumes.
func (b *FakeVolumeBinder) AssumePodVolumes(assumedPod *v1.Pod, nodeName string) (bool, error) {
b.AssumeCalled = true
return b.config.AllBound, b.config.AssumeErr
}
// BindPodVolumes implements SchedulerVolumeBinder.BindPodVolumes.
func (b *FakeVolumeBinder) BindPodVolumes(assumedPod *v1.Pod) error {
b.BindCalled = true
return b.config.BindErr
}
// GetBindingsCache implements SchedulerVolumeBinder.GetBindingsCache.
func (b *FakeVolumeBinder) GetBindingsCache() PodBindingCache {
return nil
}
// DeletePodBindings implements SchedulerVolumeBinder.DeletePodBindings.
func (b *FakeVolumeBinder) DeletePodBindings(pod *v1.Pod) {}

View File

@ -59,14 +59,6 @@ const (
// Enables support of HPA scaling to zero pods when an object or custom metric is configured.
HPAScaleToZero featuregate.Feature = "HPAScaleToZero"
// owner: @Huang-Wei
// beta: v1.13
// ga: v1.18
//
// Changes the logic behind evicting Pods from not ready Nodes
// to take advantage of NoExecute Taints and Tolerations.
TaintBasedEvictions featuregate.Feature = "TaintBasedEvictions"
// owner: @mikedanese
// alpha: v1.7
// beta: v1.12
@ -78,6 +70,7 @@ const (
// owner: @mikedanese
// beta: v1.8
// ga: v1.19
//
// Automatically renews the client certificate used for communicating with
// the API server as the certificate approaches expiration.
@ -111,14 +104,6 @@ const (
// Allows running an ephemeral container in pod namespaces to troubleshoot a running pod.
EphemeralContainers featuregate.Feature = "EphemeralContainers"
// owner: @verb
// alpha: v1.10
// beta: v1.12
// GA: v1.17
//
// Allows all containers in a pod to share a process namespace.
PodShareProcessNamespace featuregate.Feature = "PodShareProcessNamespace"
// owner: @sjenning
// alpha: v1.11
//
@ -154,18 +139,21 @@ const (
// owner @smarterclayton
// alpha: v1.16
// beta: v1.19
//
// Enable legacy behavior to vary cluster functionality on the node-role.kubernetes.io labels. On by default (legacy), will be turned off in 1.18.
LegacyNodeRoleBehavior featuregate.Feature = "LegacyNodeRoleBehavior"
// owner @brendandburns
// alpha: v1.9
// beta: v1.19
//
// Enable nodes to exclude themselves from service load balancers
ServiceNodeExclusion featuregate.Feature = "ServiceNodeExclusion"
// owner @smarterclayton
// alpha: v1.16
// beta: v1.19
//
// Enable nodes to exclude themselves from network disruption checks
NodeDisruptionExclusion featuregate.Feature = "NodeDisruptionExclusion"
@ -198,21 +186,10 @@ const (
// Postpone deletion of a PV or a PVC when they are being used
StorageObjectInUseProtection featuregate.Feature = "StorageObjectInUseProtection"
// owner: @aveshagarwal
// alpha: v1.9
//
// Enable resource limits priority function
ResourceLimitsPriorityFunction featuregate.Feature = "ResourceLimitsPriorityFunction"
// owner: @m1093782566
// GA: v1.11
//
// Implement IPVS-based in-cluster service load balancing
SupportIPVSProxyMode featuregate.Feature = "SupportIPVSProxyMode"
// owner: @dims, @derekwaynecarr
// alpha: v1.10
// beta: v1.14
// GA: v1.20
//
// Implement support for limiting pids in pods
SupportPodPidsLimit featuregate.Feature = "SupportPodPidsLimit"
@ -221,16 +198,19 @@ const (
// alpha: v1.10
//
// Enable Hyper-V containers on Windows
// Deprecated in 1.20 and removed in 1.21
HyperVContainer featuregate.Feature = "HyperVContainer"
// owner: @mikedanese
// beta: v1.12
// ga: v1.20
//
// Implement TokenRequest endpoint on service account resources.
TokenRequest featuregate.Feature = "TokenRequest"
// owner: @mikedanese
// beta: v1.12
// ga: v1.20
//
// Enable ServiceAccountTokenVolumeProjection support in ProjectedVolumes.
TokenRequestProjection featuregate.Feature = "TokenRequestProjection"
@ -245,6 +225,7 @@ const (
// owner: @mtaufen
// alpha: v1.18
// beta: v1.20
//
// Enable OIDC discovery endpoints (issuer and JWKS URLs) for the service
// account issuer in the API server.
@ -288,15 +269,6 @@ const (
// while making decisions.
BalanceAttachedNodeVolumes featuregate.Feature = "BalanceAttachedNodeVolumes"
// owner: @kevtaylor
// alpha: v1.14
// beta: v1.15
// ga: v1.17
//
// Allow subpath environment variable substitution
// Only applicable if the VolumeSubpath feature is also enabled
VolumeSubpathEnvExpansion featuregate.Feature = "VolumeSubpathEnvExpansion"
// owner: @vladimirvivien
// alpha: v1.11
// beta: v1.14
@ -305,16 +277,36 @@ const (
// Enables CSI to use raw block storage volumes
CSIBlockVolume featuregate.Feature = "CSIBlockVolume"
// owner: @vladimirvivien
// owner: @pohly
// alpha: v1.14
// beta: v1.16
//
// Enables CSI Inline volumes support for pods
CSIInlineVolume featuregate.Feature = "CSIInlineVolume"
// owner: @pohly
// alpha: v1.19
//
// Enables tracking of available storage capacity that CSI drivers provide.
CSIStorageCapacity featuregate.Feature = "CSIStorageCapacity"
// owner: @alculquicondor
// beta: v1.20
//
// Enables the use of PodTopologySpread scheduling plugin to do default
// spreading and disables legacy SelectorSpread plugin.
DefaultPodTopologySpread featuregate.Feature = "DefaultPodTopologySpread"
// owner: @pohly
// alpha: v1.19
//
// Enables generic ephemeral inline volume support for pods
GenericEphemeralVolume featuregate.Feature = "GenericEphemeralVolume"
// owner: @tallclair
// alpha: v1.12
// beta: v1.14
// GA: v1.20
//
// Enables RuntimeClass, for selecting between multiple runtimes to run a pod.
RuntimeClass featuregate.Feature = "RuntimeClass"
@ -330,6 +322,8 @@ const (
// owner: @janosi
// alpha: v1.12
// beta: v1.18
// GA: v1.20
//
// Enables SCTP as new protocol for Service ports, NetworkPolicy, and ContainerPort in Pod/Containers definition
SCTPSupport featuregate.Feature = "SCTPSupport"
@ -337,6 +331,7 @@ const (
// owner: @xing-yang
// alpha: v1.12
// beta: v1.17
// GA: v1.20
//
// Enable volume snapshot data source support.
VolumeSnapshotDataSource featuregate.Feature = "VolumeSnapshotDataSource"
@ -397,6 +392,7 @@ const (
// owner: @andyzhangx
// alpha: v1.15
// beta: v1.19
//
// Enables the Azure Disk in-tree driver to Azure Disk Driver migration feature.
CSIMigrationAzureDisk featuregate.Feature = "CSIMigrationAzureDisk"
@ -421,14 +417,36 @@ const (
// Expects Azure File CSI Driver to be installed and configured on all nodes.
CSIMigrationAzureFileComplete featuregate.Feature = "CSIMigrationAzureFileComplete"
// owner: @divyenpatel
// beta: v1.19 (requires: vSphere vCenter/ESXi Version: 7.0u1, HW Version: VM version 15)
//
// Enables the vSphere in-tree driver to vSphere CSI Driver migration feature.
CSIMigrationvSphere featuregate.Feature = "CSIMigrationvSphere"
// owner: @divyenpatel
// beta: v1.19 (requires: vSphere vCenter/ESXi Version: 7.0u1, HW Version: VM version 15)
//
// Disables the vSphere in-tree driver.
// Expects vSphere CSI Driver to be installed and configured on all nodes.
CSIMigrationvSphereComplete featuregate.Feature = "CSIMigrationvSphereComplete"
// owner: @huffmanca
// alpha: v1.19
// beta: v1.20
//
// Determines if a CSI Driver supports applying fsGroup.
CSIVolumeFSGroupPolicy featuregate.Feature = "CSIVolumeFSGroupPolicy"
// owner: @gnufied
// alpha: v1.18
// beta: v1.20
// Allows user to configure volume permission change policy for fsGroups when mounting
// a volume in a Pod.
ConfigurableFSGroupPolicy featuregate.Feature = "ConfigurableFSGroupPolicy"
// owner: @RobertKrawitz
// owner: @RobertKrawitz, @derekwaynecarr
// beta: v1.15
// GA: v1.20
//
// Implement support for limiting pids in nodes
SupportNodePidsLimit featuregate.Feature = "SupportNodePidsLimit"
@ -462,14 +480,6 @@ const (
// Expects the OpenStack Cinder CSI Driver to be installed and configured on all nodes.
CSIMigrationOpenStackComplete featuregate.Feature = "CSIMigrationOpenStackComplete"
// owner: @MrHohn
// alpha: v1.15
// beta: v1.16
// GA: v1.17
//
// Enables Finalizer Protection for Service LoadBalancers.
ServiceLoadBalancerFinalizer featuregate.Feature = "ServiceLoadBalancerFinalizer"
// owner: @RobertKrawitz
// alpha: v1.15
//
@ -479,6 +489,7 @@ const (
// owner: @denkensk
// alpha: v1.15
// beta: v1.19
//
// Enables NonPreempting option for priorityClass and pod.
NonPreemptingPriority featuregate.Feature = "NonPreemptingPriority"
@ -512,12 +523,21 @@ const (
// owner: @robscott @freehan
// alpha: v1.18
// beta: v1.19
//
// Enable Endpoint Slice consumption by kube-proxy for improved scalability.
EndpointSliceProxying featuregate.Feature = "EndpointSliceProxying"
// owner: @robscott @kumarvin123
// alpha: v1.19
//
// Enable Endpoint Slice consumption by kube-proxy in Windows for improved scalability.
WindowsEndpointSliceProxying featuregate.Feature = "WindowsEndpointSliceProxying"
// owner: @Huang-Wei
// alpha: v1.16
// beta: v1.18
// GA: v1.19
//
// Schedule pods evenly across available topology domains.
EvenPodsSpread featuregate.Feature = "EvenPodsSpread"
@ -525,6 +545,7 @@ const (
// owner: @matthyx
// alpha: v1.16
// beta: v1.18
// GA: v1.20
//
// Enables the startupProbe in kubelet worker.
StartupProbe featuregate.Feature = "StartupProbe"
@ -542,6 +563,16 @@ const (
// Enable all logic related to the PodDisruptionBudget API object in policy
PodDisruptionBudget featuregate.Feature = "PodDisruptionBudget"
// owner: @alaypatel07, @soltysh
// alpha: v1.20
// beta: v1.21
//
// CronJobControllerV2 controls whether the controller manager starts old cronjob
// controller or new one which is implemented with informers and delaying queue
//
// This feature is deprecated, and will be removed in v1.22.
CronJobControllerV2 featuregate.Feature = "CronJobControllerV2"
// owner: @m1093782566
// alpha: v1.17
//
@ -550,30 +581,34 @@ const (
// owner: @robscott
// alpha: v1.18
// beta: v1.19
// ga: v1.20
//
// Enables AppProtocol field for Services and Endpoints.
ServiceAppProtocol featuregate.Feature = "ServiceAppProtocol"
// owner: @wojtek-t
// alpha: v1.18
// beta: v1.19
//
// Enables a feature to make secrets and configmaps data immutable.
ImmutableEphemeralVolumes featuregate.Feature = "ImmutableEphemeralVolumes"
// owner: @robscott
// beta: v1.18
//
// Enables DefaultIngressClass admission controller.
DefaultIngressClass featuregate.Feature = "DefaultIngressClass"
// owner: @bart0sh
// alpha: v1.18
// beta: v1.19
//
// Enables usage of HugePages-<size> in a volume medium,
// e.g. emptyDir:
// medium: HugePages-1Gi
HugePageStorageMediumSize featuregate.Feature = "HugePageStorageMediumSize"
// owner: @derekwaynecarr
// alpha: v1.20
//
// Enables usage of hugepages-<size> in downward API.
DownwardAPIHugePages featuregate.Feature = "DownwardAPIHugePages"
// owner: @freehan
// GA: v1.18
//
@ -586,6 +621,105 @@ const (
//
// Enables usage of any object for volume data source in PVCs
AnyVolumeDataSource featuregate.Feature = "AnyVolumeDataSource"
// owner: @javidiaz
// alpha: v1.19
// beta: v1.20
//
// Allow setting the Fully Qualified Domain Name (FQDN) in the hostname of a Pod. If a Pod does not
// have FQDN, this feature has no effect.
SetHostnameAsFQDN featuregate.Feature = "SetHostnameAsFQDN"
// owner: @ksubrmnn
// alpha: v1.14
// beta: v1.20
//
// Allows kube-proxy to run in Overlay mode for Windows
WinOverlay featuregate.Feature = "WinOverlay"
// owner: @ksubrmnn
// alpha: v1.14
//
// Allows kube-proxy to create DSR loadbalancers for Windows
WinDSR featuregate.Feature = "WinDSR"
// owner: @RenaudWasTaken @dashpole
// alpha: v1.19
// beta: v1.20
//
// Disables Accelerator Metrics Collected by Kubelet
DisableAcceleratorUsageMetrics featuregate.Feature = "DisableAcceleratorUsageMetrics"
// owner: @arjunrn @mwielgus @josephburnett
// alpha: v1.20
//
// Add support for the HPA to scale based on metrics from individual containers
// in target pods
HPAContainerMetrics featuregate.Feature = "HPAContainerMetrics"
// owner: @zshihang
// alpha: v1.13
// beta: v1.20
//
// Allows kube-controller-manager to publish kube-root-ca.crt configmap to
// every namespace. This feature is a prerequisite of BoundServiceAccountTokenVolume.
RootCAConfigMap featuregate.Feature = "RootCAConfigMap"
// owner: @andrewsykim
// alpha: v1.20
//
// Enable Terminating condition in Endpoint Slices.
EndpointSliceTerminatingCondition featuregate.Feature = "EndpointSliceTerminatingCondition"
// owner: @robscott
// alpha: v1.20
//
// Enable NodeName field on Endpoint Slices.
EndpointSliceNodeName featuregate.Feature = "EndpointSliceNodeName"
// owner: @derekwaynecarr
// alpha: v1.20
//
// Enables kubelet support to size memory backed volumes
SizeMemoryBackedVolumes featuregate.Feature = "SizeMemoryBackedVolumes"
// owner: @andrewsykim @SergeyKanzhelev
// GA: v1.20
//
// Ensure kubelet respects exec probe timeouts. Feature gate exists in-case existing workloads
// may depend on old behavior where exec probe timeouts were ignored.
// Lock to default in v1.21 and remove in v1.22.
ExecProbeTimeout featuregate.Feature = "ExecProbeTimeout"
// owner: @andrewsykim
// alpha: v1.20
//
// Enable kubelet exec plugins for image pull credentials.
KubeletCredentialProviders featuregate.Feature = "KubeletCredentialProviders"
// owner: @zshihang
// alpha: v1.20
//
// Enable kubelet to pass pod's service account token to NodePublishVolume
// call of CSI driver which is mounting volumes for that pod.
CSIServiceAccountToken featuregate.Feature = "CSIServiceAccountToken"
// owner: @bobbypage
// alpha: v1.20
// Adds support for kubelet to detect node shutdown and gracefully terminate pods prior to the node being shutdown.
GracefulNodeShutdown featuregate.Feature = "GracefulNodeShutdown"
// owner: @andrewsykim @uablrek
// alpha: v1.20
//
// Allows control if NodePorts shall be created for services with "type: LoadBalancer" by defining the spec.AllocateLoadBalancerNodePorts field (bool)
ServiceLBNodePortControl featuregate.Feature = "ServiceLBNodePortControl"
// owner: @janosi
// alpha: v1.20
//
// Enables the usage of different protocols in the same Service with type=LoadBalancer
MixedProtocolLBService featuregate.Feature = "MixedProtocolLBService"
)
func init() {
@ -600,13 +734,11 @@ var defaultKubernetesFeatureGates = map[featuregate.Feature]featuregate.FeatureS
DynamicKubeletConfig: {Default: true, PreRelease: featuregate.Beta},
ExperimentalHostUserNamespaceDefaultingGate: {Default: false, PreRelease: featuregate.Beta},
DevicePlugins: {Default: true, PreRelease: featuregate.Beta},
TaintBasedEvictions: {Default: true, PreRelease: featuregate.GA, LockToDefault: true}, // remove in 1.19
RotateKubeletServerCertificate: {Default: true, PreRelease: featuregate.Beta},
RotateKubeletClientCertificate: {Default: true, PreRelease: featuregate.Beta},
RotateKubeletClientCertificate: {Default: true, PreRelease: featuregate.GA, LockToDefault: true}, // remove in 1.20
LocalStorageCapacityIsolation: {Default: true, PreRelease: featuregate.Beta},
Sysctls: {Default: true, PreRelease: featuregate.Beta},
EphemeralContainers: {Default: false, PreRelease: featuregate.Alpha},
PodShareProcessNamespace: {Default: true, PreRelease: featuregate.GA, LockToDefault: true}, // remove in 1.19
QOSReserved: {Default: false, PreRelease: featuregate.Alpha},
ExpandPersistentVolumes: {Default: true, PreRelease: featuregate.Beta},
ExpandInUsePersistentVolumes: {Default: true, PreRelease: featuregate.Beta},
@ -615,83 +747,102 @@ var defaultKubernetesFeatureGates = map[featuregate.Feature]featuregate.FeatureS
CPUManager: {Default: true, PreRelease: featuregate.Beta},
CPUCFSQuotaPeriod: {Default: false, PreRelease: featuregate.Alpha},
TopologyManager: {Default: true, PreRelease: featuregate.Beta},
ServiceNodeExclusion: {Default: false, PreRelease: featuregate.Alpha},
NodeDisruptionExclusion: {Default: false, PreRelease: featuregate.Alpha},
ServiceNodeExclusion: {Default: true, PreRelease: featuregate.Beta},
NodeDisruptionExclusion: {Default: true, PreRelease: featuregate.Beta},
CSIDriverRegistry: {Default: true, PreRelease: featuregate.GA, LockToDefault: true}, // remove in 1.20
CSINodeInfo: {Default: true, PreRelease: featuregate.GA, LockToDefault: true}, // remove in 1.19
BlockVolume: {Default: true, PreRelease: featuregate.GA, LockToDefault: true}, // remove in 1.20
StorageObjectInUseProtection: {Default: true, PreRelease: featuregate.GA},
ResourceLimitsPriorityFunction: {Default: false, PreRelease: featuregate.Alpha},
SupportIPVSProxyMode: {Default: true, PreRelease: featuregate.GA},
SupportPodPidsLimit: {Default: true, PreRelease: featuregate.Beta},
SupportNodePidsLimit: {Default: true, PreRelease: featuregate.Beta},
HyperVContainer: {Default: false, PreRelease: featuregate.Alpha},
TokenRequest: {Default: true, PreRelease: featuregate.Beta},
TokenRequestProjection: {Default: true, PreRelease: featuregate.Beta},
SupportPodPidsLimit: {Default: true, PreRelease: featuregate.GA, LockToDefault: true}, // remove in 1.23
SupportNodePidsLimit: {Default: true, PreRelease: featuregate.GA, LockToDefault: true}, // remove in 1.23
HyperVContainer: {Default: false, PreRelease: featuregate.Deprecated},
TokenRequest: {Default: true, PreRelease: featuregate.GA, LockToDefault: true}, // remove in 1.21
TokenRequestProjection: {Default: true, PreRelease: featuregate.GA, LockToDefault: true}, // remove in 1.21
BoundServiceAccountTokenVolume: {Default: false, PreRelease: featuregate.Alpha},
ServiceAccountIssuerDiscovery: {Default: false, PreRelease: featuregate.Alpha},
ServiceAccountIssuerDiscovery: {Default: true, PreRelease: featuregate.Beta},
CRIContainerLogRotation: {Default: true, PreRelease: featuregate.Beta},
CSIMigration: {Default: true, PreRelease: featuregate.Beta},
CSIMigrationGCE: {Default: false, PreRelease: featuregate.Beta}, // Off by default (requires GCE PD CSI Driver)
CSIMigrationGCEComplete: {Default: false, PreRelease: featuregate.Alpha},
CSIMigrationAWS: {Default: false, PreRelease: featuregate.Beta}, // Off by default (requires AWS EBS CSI driver)
CSIMigrationAWSComplete: {Default: false, PreRelease: featuregate.Alpha},
CSIMigrationAzureDisk: {Default: false, PreRelease: featuregate.Alpha},
CSIMigrationAzureDisk: {Default: false, PreRelease: featuregate.Beta}, // Off by default (requires Azure Disk CSI driver)
CSIMigrationAzureDiskComplete: {Default: false, PreRelease: featuregate.Alpha},
CSIMigrationAzureFile: {Default: false, PreRelease: featuregate.Alpha},
CSIMigrationAzureFileComplete: {Default: false, PreRelease: featuregate.Alpha},
CSIMigrationvSphere: {Default: false, PreRelease: featuregate.Beta},
CSIMigrationvSphereComplete: {Default: false, PreRelease: featuregate.Beta},
RunAsGroup: {Default: true, PreRelease: featuregate.Beta},
CSIMigrationOpenStack: {Default: false, PreRelease: featuregate.Beta}, // Off by default (requires OpenStack Cinder CSI driver)
CSIMigrationOpenStackComplete: {Default: false, PreRelease: featuregate.Alpha},
VolumeSubpath: {Default: true, PreRelease: featuregate.GA},
ConfigurableFSGroupPolicy: {Default: false, PreRelease: featuregate.Alpha},
ConfigurableFSGroupPolicy: {Default: true, PreRelease: featuregate.Beta},
BalanceAttachedNodeVolumes: {Default: false, PreRelease: featuregate.Alpha},
VolumeSubpathEnvExpansion: {Default: true, PreRelease: featuregate.GA, LockToDefault: true}, // remove in 1.19,
CSIBlockVolume: {Default: true, PreRelease: featuregate.GA, LockToDefault: true}, // remove in 1.20
CSIInlineVolume: {Default: true, PreRelease: featuregate.Beta},
RuntimeClass: {Default: true, PreRelease: featuregate.Beta},
CSIStorageCapacity: {Default: false, PreRelease: featuregate.Alpha},
CSIServiceAccountToken: {Default: false, PreRelease: featuregate.Alpha},
GenericEphemeralVolume: {Default: false, PreRelease: featuregate.Alpha},
CSIVolumeFSGroupPolicy: {Default: true, PreRelease: featuregate.Beta},
RuntimeClass: {Default: true, PreRelease: featuregate.GA, LockToDefault: true}, // remove in 1.23
NodeLease: {Default: true, PreRelease: featuregate.GA, LockToDefault: true},
SCTPSupport: {Default: false, PreRelease: featuregate.Alpha},
VolumeSnapshotDataSource: {Default: true, PreRelease: featuregate.Beta},
SCTPSupport: {Default: true, PreRelease: featuregate.GA, LockToDefault: true}, // remove in 1.22
VolumeSnapshotDataSource: {Default: true, PreRelease: featuregate.GA, LockToDefault: true}, // remove in 1.21
ProcMountType: {Default: false, PreRelease: featuregate.Alpha},
TTLAfterFinished: {Default: false, PreRelease: featuregate.Alpha},
KubeletPodResources: {Default: true, PreRelease: featuregate.Beta},
WindowsGMSA: {Default: true, PreRelease: featuregate.GA, LockToDefault: true}, // remove in 1.20
WindowsRunAsUserName: {Default: true, PreRelease: featuregate.GA, LockToDefault: true}, // remove in 1.20
ServiceLoadBalancerFinalizer: {Default: true, PreRelease: featuregate.GA, LockToDefault: true},
LocalStorageCapacityIsolationFSQuotaMonitoring: {Default: false, PreRelease: featuregate.Alpha},
NonPreemptingPriority: {Default: false, PreRelease: featuregate.Alpha},
NonPreemptingPriority: {Default: true, PreRelease: featuregate.Beta},
VolumePVCDataSource: {Default: true, PreRelease: featuregate.GA, LockToDefault: true}, // remove in 1.20
PodOverhead: {Default: true, PreRelease: featuregate.Beta},
IPv6DualStack: {Default: false, PreRelease: featuregate.Alpha},
EndpointSlice: {Default: true, PreRelease: featuregate.Beta},
EndpointSliceProxying: {Default: false, PreRelease: featuregate.Alpha},
EvenPodsSpread: {Default: true, PreRelease: featuregate.Beta},
StartupProbe: {Default: true, PreRelease: featuregate.Beta},
EndpointSliceProxying: {Default: true, PreRelease: featuregate.Beta},
EndpointSliceTerminatingCondition: {Default: false, PreRelease: featuregate.Alpha},
EndpointSliceNodeName: {Default: false, PreRelease: featuregate.Alpha},
WindowsEndpointSliceProxying: {Default: false, PreRelease: featuregate.Alpha},
EvenPodsSpread: {Default: true, PreRelease: featuregate.GA, LockToDefault: true}, // remove in 1.21
StartupProbe: {Default: true, PreRelease: featuregate.GA, LockToDefault: true}, // remove in 1.23
AllowInsecureBackendProxy: {Default: true, PreRelease: featuregate.Beta},
PodDisruptionBudget: {Default: true, PreRelease: featuregate.Beta},
CronJobControllerV2: {Default: false, PreRelease: featuregate.Alpha},
ServiceTopology: {Default: false, PreRelease: featuregate.Alpha},
ServiceAppProtocol: {Default: false, PreRelease: featuregate.Alpha},
ImmutableEphemeralVolumes: {Default: false, PreRelease: featuregate.Alpha},
DefaultIngressClass: {Default: true, PreRelease: featuregate.Beta},
HugePageStorageMediumSize: {Default: false, PreRelease: featuregate.Alpha},
ExternalPolicyForExternalIP: {Default: false, PreRelease: featuregate.GA}, // remove in 1.19
ServiceAppProtocol: {Default: true, PreRelease: featuregate.GA, LockToDefault: true},
ImmutableEphemeralVolumes: {Default: true, PreRelease: featuregate.Beta},
HugePageStorageMediumSize: {Default: true, PreRelease: featuregate.Beta},
DownwardAPIHugePages: {Default: false, PreRelease: featuregate.Alpha},
ExternalPolicyForExternalIP: {Default: true, PreRelease: featuregate.GA, LockToDefault: true}, // remove in 1.22
AnyVolumeDataSource: {Default: false, PreRelease: featuregate.Alpha},
DefaultPodTopologySpread: {Default: true, PreRelease: featuregate.Beta},
SetHostnameAsFQDN: {Default: true, PreRelease: featuregate.Beta},
WinOverlay: {Default: true, PreRelease: featuregate.Beta},
WinDSR: {Default: false, PreRelease: featuregate.Alpha},
DisableAcceleratorUsageMetrics: {Default: true, PreRelease: featuregate.Beta},
HPAContainerMetrics: {Default: false, PreRelease: featuregate.Alpha},
RootCAConfigMap: {Default: true, PreRelease: featuregate.Beta},
SizeMemoryBackedVolumes: {Default: false, PreRelease: featuregate.Alpha},
ExecProbeTimeout: {Default: true, PreRelease: featuregate.GA}, // lock to default in v1.21 and remove in v1.22
KubeletCredentialProviders: {Default: false, PreRelease: featuregate.Alpha},
GracefulNodeShutdown: {Default: false, PreRelease: featuregate.Alpha},
ServiceLBNodePortControl: {Default: false, PreRelease: featuregate.Alpha},
MixedProtocolLBService: {Default: false, PreRelease: featuregate.Alpha},
// inherited features from generic apiserver, relisted here to get a conflict if it is changed
// unintentionally on either side:
genericfeatures.StreamingProxyRedirects: {Default: true, PreRelease: featuregate.Deprecated},
genericfeatures.ValidateProxyRedirects: {Default: true, PreRelease: featuregate.Beta},
genericfeatures.AdvancedAuditing: {Default: true, PreRelease: featuregate.GA},
genericfeatures.DynamicAuditing: {Default: false, PreRelease: featuregate.Alpha},
genericfeatures.APIResponseCompression: {Default: true, PreRelease: featuregate.Beta},
genericfeatures.APIListChunking: {Default: true, PreRelease: featuregate.Beta},
genericfeatures.DryRun: {Default: true, PreRelease: featuregate.Beta},
genericfeatures.DryRun: {Default: true, PreRelease: featuregate.GA},
genericfeatures.ServerSideApply: {Default: true, PreRelease: featuregate.Beta},
genericfeatures.APIPriorityAndFairness: {Default: false, PreRelease: featuregate.Alpha},
genericfeatures.APIPriorityAndFairness: {Default: true, PreRelease: featuregate.Beta},
genericfeatures.WarningHeaders: {Default: true, PreRelease: featuregate.Beta},
// features that enable backwards compatibility but are scheduled to be removed
// ...
HPAScaleToZero: {Default: false, PreRelease: featuregate.Alpha},
LegacyNodeRoleBehavior: {Default: true, PreRelease: featuregate.Alpha},
LegacyNodeRoleBehavior: {Default: true, PreRelease: featuregate.Beta},
}

View File

@ -1,45 +0,0 @@
package(default_visibility = ["//visibility:public"])
load(
"@io_bazel_rules_go//go:def.bzl",
"go_library",
)
go_library(
name = "go_default_library",
srcs = [
"well_known_annotations.go",
"well_known_annotations_windows.go",
"well_known_labels.go",
],
importpath = "k8s.io/kubernetes/pkg/kubelet/apis",
deps = [
"//staging/src/k8s.io/api/core/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/sets:go_default_library",
] + select({
"@io_bazel_rules_go//go/platform:windows": [
"//pkg/features:go_default_library",
"//staging/src/k8s.io/apiserver/pkg/util/feature:go_default_library",
],
"//conditions:default": [],
}),
)
filegroup(
name = "package-srcs",
srcs = glob(["**"]),
tags = ["automanaged"],
visibility = ["//visibility:private"],
)
filegroup(
name = "all-srcs",
srcs = [
":package-srcs",
"//pkg/kubelet/apis/config:all-srcs",
"//pkg/kubelet/apis/podresources:all-srcs",
"//pkg/kubelet/apis/resourcemetrics/v1alpha1:all-srcs",
"//pkg/kubelet/apis/stats/v1alpha1:all-srcs",
],
tags = ["automanaged"],
)

View File

@ -21,6 +21,7 @@ go_library(
"//staging/src/k8s.io/apimachinery/pkg/apis/meta/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/runtime:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/runtime/schema:go_default_library",
"//staging/src/k8s.io/component-base/config:go_default_library",
],
)
@ -37,6 +38,7 @@ filegroup(
":package-srcs",
"//pkg/kubelet/apis/config/fuzzer:all-srcs",
"//pkg/kubelet/apis/config/scheme:all-srcs",
"//pkg/kubelet/apis/config/v1alpha1:all-srcs",
"//pkg/kubelet/apis/config/v1beta1:all-srcs",
"//pkg/kubelet/apis/config/validation:all-srcs",
],
@ -45,10 +47,16 @@ filegroup(
go_test(
name = "go_default_test",
srcs = ["helpers_test.go"],
srcs = [
"helpers_test.go",
"register_test.go",
],
embed = [":go_default_library"],
deps = [
"//staging/src/k8s.io/api/core/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/apis/meta/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/sets:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/validation/field:go_default_library",
"//staging/src/k8s.io/component-base/config/testing:go_default_library",
],
)

View File

@ -1,6 +1,9 @@
# See the OWNERS docs at https://go.k8s.io/owners
# Disable inheritance as this is an api owners file
options:
no_parent_owners: true
approvers:
- mtaufen
- api-approvers
reviewers:
- sig-node-reviewers
- sig-node-api-reviewers

View File

@ -26,5 +26,6 @@ func KubeletConfigurationPathRefs(kc *KubeletConfiguration) []*string {
paths = append(paths, &kc.TLSCertFile)
paths = append(paths, &kc.TLSPrivateKeyFile)
paths = append(paths, &kc.ResolverConfig)
paths = append(paths, &kc.VolumePluginDir)
return paths
}

View File

@ -39,6 +39,7 @@ func addKnownTypes(scheme *runtime.Scheme) error {
scheme.AddKnownTypes(SchemeGroupVersion,
&KubeletConfiguration{},
&SerializedNodeConfigSource{},
&CredentialProviderConfig{},
)
return nil
}

View File

@ -19,6 +19,7 @@ package config
import (
v1 "k8s.io/api/core/v1"
metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
componentbaseconfig "k8s.io/component-base/config"
)
// HairpinMode denotes how the kubelet should configure networking to handle
@ -60,12 +61,18 @@ const (
// BestEffortTopologyManagerPolicy is a mode in which kubelet will favour
// pods with NUMA alignment of CPU and device resources.
BestEffortTopologyManagerPolicy = "best-effort"
// NoneTopologyManager Policy is a mode in which kubelet has no knowledge
// NoneTopologyManagerPolicy is a mode in which kubelet has no knowledge
// of NUMA alignment of a pod's CPU and device resources.
NoneTopologyManagerPolicy = "none"
// SingleNumaNodeTopologyManager Policy iis a mode in which kubelet only allows
// SingleNumaNodeTopologyManagerPolicy is a mode in which kubelet only allows
// pods with a single NUMA alignment of CPU and device resources.
SingleNumaNodeTopologyManager = "single-numa-node"
SingleNumaNodeTopologyManagerPolicy = "single-numa-node"
// ContainerTopologyManagerScope represents that
// topology policy is applied on a per-container basis.
ContainerTopologyManagerScope = "container"
// PodTopologyManagerScope represents that
// topology policy is applied on a per-pod basis.
PodTopologyManagerScope = "pod"
)
// +k8s:deepcopy-gen:interfaces=k8s.io/apimachinery/pkg/runtime.Object
@ -74,6 +81,9 @@ const (
type KubeletConfiguration struct {
metav1.TypeMeta
// enableServer enables Kubelet's secured server.
// Note: Kubelet's insecure port is controlled by the readOnlyPort option.
EnableServer bool
// staticPodPath is the path to the directory containing local (static) pods to
// run, or the path to a single static pod file.
StaticPodPath string
@ -97,6 +107,12 @@ type KubeletConfiguration struct {
// readOnlyPort is the read-only port for the Kubelet to serve on with
// no authentication/authorization (set to 0 to disable)
ReadOnlyPort int32
// volumePluginDir is the full path of the directory in which to search
// for additional third party volume plugins.
VolumePluginDir string
// providerID, if set, sets the unique id of the instance that an external provider (i.e. cloudprovider)
// can use to identify a specific node
ProviderID string
// tlsCertFile is the file containing x509 Certificate for HTTPS. (CA cert,
// if any, concatenated after server cert). If tlsCertFile and
// tlsPrivateKeyFile are not provided, a self-signed certificate
@ -113,8 +129,7 @@ type KubeletConfiguration struct {
TLSMinVersion string
// rotateCertificates enables client certificate rotation. The Kubelet will request a
// new certificate from the certificates.k8s.io API. This requires an approver to approve the
// certificate signing requests. The RotateKubeletClientCertificate feature
// must be enabled.
// certificate signing requests.
RotateCertificates bool
// serverTLSBootstrap enables server certificate bootstrap. Instead of self
// signing a serving certificate, the Kubelet will request a certificate from
@ -212,6 +227,12 @@ type KubeletConfiguration struct {
// TopologyManagerPolicy is the name of the policy to use.
// Policies other than "none" require the TopologyManager feature gate to be enabled.
TopologyManagerPolicy string
// TopologyManagerScope represents the scope of topology hint generation
// that topology manager requests and hint providers generate.
// "pod" scope requires the TopologyManager feature gate to be enabled.
// Default: "container"
// +optional
TopologyManagerScope string
// Map of QoS resource reservation percentages (memory only for now).
// Requires the QOSReserved feature gate to be enabled.
QOSReserved map[string]string
@ -238,6 +259,9 @@ type KubeletConfiguration struct {
// ResolverConfig is the resolver configuration file used as the basis
// for the container DNS resolution configuration.
ResolverConfig string
// RunOnce causes the Kubelet to check the API server once for pods,
// run those in addition to the pods specified by static pod files, and exit.
RunOnce bool
// cpuCFSQuota enables CPU CFS quota enforcement for containers that
// specify CPU limits
CPUCFSQuota bool
@ -245,6 +269,8 @@ type KubeletConfiguration struct {
CPUCFSQuotaPeriod metav1.Duration
// maxOpenFiles is Number of files that can be opened by Kubelet process.
MaxOpenFiles int64
// nodeStatusMaxImages caps the number of images reported in Node.Status.Images.
NodeStatusMaxImages int32
// contentType is contentType of requests sent to apiserver.
ContentType string
// kubeAPIQPS is the QPS to use while talking with kubernetes apiserver
@ -308,6 +334,9 @@ type KubeletConfiguration struct {
// These sysctls are namespaced but not allowed by default. For example: "kernel.msg*,net.ipv4.route.min_pmtu"
// +optional
AllowedUnsafeSysctls []string
// kernelMemcgNotification if enabled, the kubelet will integrate with the kernel memcg
// notification to determine if memory eviction thresholds are crossed rather than polling.
KernelMemcgNotification bool
/* the following fields are meant for Node Allocatable */
@ -341,6 +370,18 @@ type KubeletConfiguration struct {
// The purpose of this format is make sure you have the opportunity to notice if the next release hides additional metrics,
// rather than being surprised when they are permanently removed in the release after that.
ShowHiddenMetricsForVersion string
// Logging specifies the options of logging.
// Refer [Logs Options](https://github.com/kubernetes/component-base/blob/master/logs/options.go) for more information.
Logging componentbaseconfig.LoggingConfiguration
// EnableSystemLogHandler enables /logs handler.
EnableSystemLogHandler bool
// ShutdownGracePeriod specifies the total duration that the node should delay the shutdown and total grace period for pod termination during a node shutdown.
// Defaults to 30 seconds, requires GracefulNodeShutdown feature gate to be enabled.
ShutdownGracePeriod metav1.Duration
// ShutdownGracePeriodCriticalPods specifies the duration used to terminate critical pods during a node shutdown. This should be less than ShutdownGracePeriod.
// Defaults to 10 seconds, requires GracefulNodeShutdown feature gate to be enabled.
// For example, if ShutdownGracePeriod=30s, and ShutdownGracePeriodCriticalPods=10s, during a node shutdown the first 20 seconds would be reserved for gracefully terminating normal pods, and the last 10 seconds would be reserved for terminating critical pods.
ShutdownGracePeriodCriticalPods metav1.Duration
}
// KubeletAuthorizationMode denotes the authorization mode for the kubelet
@ -399,7 +440,7 @@ type KubeletWebhookAuthentication struct {
CacheTTL metav1.Duration
}
// KubeletAnonymousAuthentication enables anonymous requests to the kubetlet server.
// KubeletAnonymousAuthentication enables anonymous requests to the kubelet server.
type KubeletAnonymousAuthentication struct {
// enabled allows anonymous requests to the kubelet server.
// Requests that are not rejected by another authentication method are treated as anonymous requests.
@ -418,3 +459,79 @@ type SerializedNodeConfigSource struct {
// +optional
Source v1.NodeConfigSource
}
// +k8s:deepcopy-gen:interfaces=k8s.io/apimachinery/pkg/runtime.Object
// CredentialProviderConfig is the configuration containing information about
// each exec credential provider. Kubelet reads this configuration from disk and enables
// each provider as specified by the CredentialProvider type.
type CredentialProviderConfig struct {
metav1.TypeMeta
// providers is a list of credential provider plugins that will be enabled by the kubelet.
// Multiple providers may match against a single image, in which case credentials
// from all providers will be returned to the kubelet. If multiple providers are called
// for a single image, the results are combined. If providers return overlapping
// auth keys, the value from the provider earlier in this list is used.
Providers []CredentialProvider
}
// CredentialProvider represents an exec plugin to be invoked by the kubelet. The plugin is only
// invoked when an image being pulled matches the images handled by the plugin (see matchImages).
type CredentialProvider struct {
// name is the required name of the credential provider. It must match the name of the
// provider executable as seen by the kubelet. The executable must be in the kubelet's
// bin directory (set by the --credential-provider-bin-dir flag).
Name string
// matchImages is a required list of strings used to match against images in order to
// determine if this provider should be invoked. If one of the strings matches the
// requested image from the kubelet, the plugin will be invoked and given a chance
// to provide credentials. Images are expected to contain the registry domain
// and URL path.
//
// Each entry in matchImages is a pattern which can optionally contain a port and a path.
// Globs can be used in the domain, but not in the port or the path. Globs are supported
// as subdomains like '*.k8s.io' or 'k8s.*.io', and top-level-domains such as 'k8s.*'.
// Matching partial subdomains like 'app*.k8s.io' is also supported. Each glob can only match
// a single subdomain segment, so *.io does not match *.k8s.io.
//
// A match exists between an image and a matchImage when all of the below are true:
// - Both contain the same number of domain parts and each part matches.
// - The URL path of an imageMatch must be a prefix of the target image URL path.
// - If the imageMatch contains a port, then the port must match in the image as well.
//
// Example values of matchImages:
// - 123456789.dkr.ecr.us-east-1.amazonaws.com
// - *.azurecr.io
// - gcr.io
// - *.*.registry.io
// - registry.io:8080/path
MatchImages []string
// defaultCacheDuration is the default duration the plugin will cache credentials in-memory
// if a cache duration is not provided in the plugin response. This field is required.
DefaultCacheDuration *metav1.Duration
// Required input version of the exec CredentialProviderRequest. The returned CredentialProviderResponse
// MUST use the same encoding version as the input. Current supported values are:
// - credentialprovider.kubelet.k8s.io/v1alpha1
APIVersion string
// Arguments to pass to the command when executing it.
// +optional
Args []string
// Env defines additional environment variables to expose to the process. These
// are unioned with the host's environment, as well as variables client-go uses
// to pass argument to the plugin.
// +optional
Env []ExecEnvVar
}
// ExecEnvVar is used for setting environment variables when executing an exec-based
// credential plugin.
type ExecEnvVar struct {
Name string
Value string
}

View File

@ -21,9 +21,94 @@ limitations under the License.
package config
import (
v1 "k8s.io/apimachinery/pkg/apis/meta/v1"
runtime "k8s.io/apimachinery/pkg/runtime"
)
// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
func (in *CredentialProvider) DeepCopyInto(out *CredentialProvider) {
*out = *in
if in.MatchImages != nil {
in, out := &in.MatchImages, &out.MatchImages
*out = make([]string, len(*in))
copy(*out, *in)
}
if in.DefaultCacheDuration != nil {
in, out := &in.DefaultCacheDuration, &out.DefaultCacheDuration
*out = new(v1.Duration)
**out = **in
}
if in.Args != nil {
in, out := &in.Args, &out.Args
*out = make([]string, len(*in))
copy(*out, *in)
}
if in.Env != nil {
in, out := &in.Env, &out.Env
*out = make([]ExecEnvVar, len(*in))
copy(*out, *in)
}
return
}
// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new CredentialProvider.
func (in *CredentialProvider) DeepCopy() *CredentialProvider {
if in == nil {
return nil
}
out := new(CredentialProvider)
in.DeepCopyInto(out)
return out
}
// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
func (in *CredentialProviderConfig) DeepCopyInto(out *CredentialProviderConfig) {
*out = *in
out.TypeMeta = in.TypeMeta
if in.Providers != nil {
in, out := &in.Providers, &out.Providers
*out = make([]CredentialProvider, len(*in))
for i := range *in {
(*in)[i].DeepCopyInto(&(*out)[i])
}
}
return
}
// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new CredentialProviderConfig.
func (in *CredentialProviderConfig) DeepCopy() *CredentialProviderConfig {
if in == nil {
return nil
}
out := new(CredentialProviderConfig)
in.DeepCopyInto(out)
return out
}
// DeepCopyObject is an autogenerated deepcopy function, copying the receiver, creating a new runtime.Object.
func (in *CredentialProviderConfig) DeepCopyObject() runtime.Object {
if c := in.DeepCopy(); c != nil {
return c
}
return nil
}
// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
func (in *ExecEnvVar) DeepCopyInto(out *ExecEnvVar) {
*out = *in
return
}
// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new ExecEnvVar.
func (in *ExecEnvVar) DeepCopy() *ExecEnvVar {
if in == nil {
return nil
}
out := new(ExecEnvVar)
in.DeepCopyInto(out)
return out
}
// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
func (in *KubeletAnonymousAuthentication) DeepCopyInto(out *KubeletAnonymousAuthentication) {
*out = *in
@ -185,6 +270,9 @@ func (in *KubeletConfiguration) DeepCopyInto(out *KubeletConfiguration) {
*out = make([]string, len(*in))
copy(*out, *in)
}
out.Logging = in.Logging
out.ShutdownGracePeriod = in.ShutdownGracePeriod
out.ShutdownGracePeriodCriticalPods = in.ShutdownGracePeriodCriticalPods
return
}

View File

@ -1,26 +0,0 @@
package(default_visibility = ["//visibility:public"])
load(
"@io_bazel_rules_go//go:def.bzl",
"go_library",
)
go_library(
name = "go_default_library",
srcs = ["types.go"],
importpath = "k8s.io/kubernetes/pkg/kubelet/apis/stats/v1alpha1",
deps = ["//staging/src/k8s.io/apimachinery/pkg/apis/meta/v1:go_default_library"],
)
filegroup(
name = "package-srcs",
srcs = glob(["**"]),
tags = ["automanaged"],
visibility = ["//visibility:private"],
)
filegroup(
name = "all-srcs",
srcs = [":package-srcs"],
tags = ["automanaged"],
)

View File

@ -1,335 +0,0 @@
/*
Copyright 2015 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 v1alpha1
import (
metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
)
// Summary is a top-level container for holding NodeStats and PodStats.
type Summary struct {
// Overall node stats.
Node NodeStats `json:"node"`
// Per-pod stats.
Pods []PodStats `json:"pods"`
}
// NodeStats holds node-level unprocessed sample stats.
type NodeStats struct {
// Reference to the measured Node.
NodeName string `json:"nodeName"`
// Stats of system daemons tracked as raw containers.
// The system containers are named according to the SystemContainer* constants.
// +optional
// +patchMergeKey=name
// +patchStrategy=merge
SystemContainers []ContainerStats `json:"systemContainers,omitempty" patchStrategy:"merge" patchMergeKey:"name"`
// The time at which data collection for the node-scoped (i.e. aggregate) stats was (re)started.
StartTime metav1.Time `json:"startTime"`
// Stats pertaining to CPU resources.
// +optional
CPU *CPUStats `json:"cpu,omitempty"`
// Stats pertaining to memory (RAM) resources.
// +optional
Memory *MemoryStats `json:"memory,omitempty"`
// Stats pertaining to network resources.
// +optional
Network *NetworkStats `json:"network,omitempty"`
// Stats pertaining to total usage of filesystem resources on the rootfs used by node k8s components.
// NodeFs.Used is the total bytes used on the filesystem.
// +optional
Fs *FsStats `json:"fs,omitempty"`
// Stats about the underlying container runtime.
// +optional
Runtime *RuntimeStats `json:"runtime,omitempty"`
// Stats about the rlimit of system.
// +optional
Rlimit *RlimitStats `json:"rlimit,omitempty"`
}
// RlimitStats are stats rlimit of OS.
type RlimitStats struct {
Time metav1.Time `json:"time"`
// The max PID of OS.
MaxPID *int64 `json:"maxpid,omitempty"`
// The number of running process in the OS.
NumOfRunningProcesses *int64 `json:"curproc,omitempty"`
}
// RuntimeStats are stats pertaining to the underlying container runtime.
type RuntimeStats struct {
// Stats about the underlying filesystem where container images are stored.
// This filesystem could be the same as the primary (root) filesystem.
// Usage here refers to the total number of bytes occupied by images on the filesystem.
// +optional
ImageFs *FsStats `json:"imageFs,omitempty"`
}
const (
// SystemContainerKubelet is the container name for the system container tracking Kubelet usage.
SystemContainerKubelet = "kubelet"
// SystemContainerRuntime is the container name for the system container tracking the runtime (e.g. docker) usage.
SystemContainerRuntime = "runtime"
// SystemContainerMisc is the container name for the system container tracking non-kubernetes processes.
SystemContainerMisc = "misc"
// SystemContainerPods is the container name for the system container tracking user pods.
SystemContainerPods = "pods"
)
// PodStats holds pod-level unprocessed sample stats.
type PodStats struct {
// Reference to the measured Pod.
PodRef PodReference `json:"podRef"`
// The time at which data collection for the pod-scoped (e.g. network) stats was (re)started.
StartTime metav1.Time `json:"startTime"`
// Stats of containers in the measured pod.
// +patchMergeKey=name
// +patchStrategy=merge
Containers []ContainerStats `json:"containers" patchStrategy:"merge" patchMergeKey:"name"`
// Stats pertaining to CPU resources consumed by pod cgroup (which includes all containers' resource usage and pod overhead).
// +optional
CPU *CPUStats `json:"cpu,omitempty"`
// Stats pertaining to memory (RAM) resources consumed by pod cgroup (which includes all containers' resource usage and pod overhead).
// +optional
Memory *MemoryStats `json:"memory,omitempty"`
// Stats pertaining to network resources.
// +optional
Network *NetworkStats `json:"network,omitempty"`
// Stats pertaining to volume usage of filesystem resources.
// VolumeStats.UsedBytes is the number of bytes used by the Volume
// +optional
// +patchMergeKey=name
// +patchStrategy=merge
VolumeStats []VolumeStats `json:"volume,omitempty" patchStrategy:"merge" patchMergeKey:"name"`
// EphemeralStorage reports the total filesystem usage for the containers and emptyDir-backed volumes in the measured Pod.
// +optional
EphemeralStorage *FsStats `json:"ephemeral-storage,omitempty"`
}
// ContainerStats holds container-level unprocessed sample stats.
type ContainerStats struct {
// Reference to the measured container.
Name string `json:"name"`
// The time at which data collection for this container was (re)started.
StartTime metav1.Time `json:"startTime"`
// Stats pertaining to CPU resources.
// +optional
CPU *CPUStats `json:"cpu,omitempty"`
// Stats pertaining to memory (RAM) resources.
// +optional
Memory *MemoryStats `json:"memory,omitempty"`
// Metrics for Accelerators. Each Accelerator corresponds to one element in the array.
Accelerators []AcceleratorStats `json:"accelerators,omitempty"`
// Stats pertaining to container rootfs usage of filesystem resources.
// Rootfs.UsedBytes is the number of bytes used for the container write layer.
// +optional
Rootfs *FsStats `json:"rootfs,omitempty"`
// Stats pertaining to container logs usage of filesystem resources.
// Logs.UsedBytes is the number of bytes used for the container logs.
// +optional
Logs *FsStats `json:"logs,omitempty"`
// User defined metrics that are exposed by containers in the pod. Typically, we expect only one container in the pod to be exposing user defined metrics. In the event of multiple containers exposing metrics, they will be combined here.
// +patchMergeKey=name
// +patchStrategy=merge
UserDefinedMetrics []UserDefinedMetric `json:"userDefinedMetrics,omitempty" patchStrategy:"merge" patchMergeKey:"name"`
}
// PodReference contains enough information to locate the referenced pod.
type PodReference struct {
Name string `json:"name"`
Namespace string `json:"namespace"`
UID string `json:"uid"`
}
// InterfaceStats contains resource value data about interface.
type InterfaceStats struct {
// The name of the interface
Name string `json:"name"`
// Cumulative count of bytes received.
// +optional
RxBytes *uint64 `json:"rxBytes,omitempty"`
// Cumulative count of receive errors encountered.
// +optional
RxErrors *uint64 `json:"rxErrors,omitempty"`
// Cumulative count of bytes transmitted.
// +optional
TxBytes *uint64 `json:"txBytes,omitempty"`
// Cumulative count of transmit errors encountered.
// +optional
TxErrors *uint64 `json:"txErrors,omitempty"`
}
// NetworkStats contains data about network resources.
type NetworkStats struct {
// The time at which these stats were updated.
Time metav1.Time `json:"time"`
// Stats for the default interface, if found
InterfaceStats `json:",inline"`
Interfaces []InterfaceStats `json:"interfaces,omitempty"`
}
// CPUStats contains data about CPU usage.
type CPUStats struct {
// The time at which these stats were updated.
Time metav1.Time `json:"time"`
// Total CPU usage (sum of all cores) averaged over the sample window.
// The "core" unit can be interpreted as CPU core-nanoseconds per second.
// +optional
UsageNanoCores *uint64 `json:"usageNanoCores,omitempty"`
// Cumulative CPU usage (sum of all cores) since object creation.
// +optional
UsageCoreNanoSeconds *uint64 `json:"usageCoreNanoSeconds,omitempty"`
}
// MemoryStats contains data about memory usage.
type MemoryStats struct {
// The time at which these stats were updated.
Time metav1.Time `json:"time"`
// Available memory for use. This is defined as the memory limit - workingSetBytes.
// If memory limit is undefined, the available bytes is omitted.
// +optional
AvailableBytes *uint64 `json:"availableBytes,omitempty"`
// Total memory in use. This includes all memory regardless of when it was accessed.
// +optional
UsageBytes *uint64 `json:"usageBytes,omitempty"`
// The amount of working set memory. This includes recently accessed memory,
// dirty memory, and kernel memory. WorkingSetBytes is <= UsageBytes
// +optional
WorkingSetBytes *uint64 `json:"workingSetBytes,omitempty"`
// The amount of anonymous and swap cache memory (includes transparent
// hugepages).
// +optional
RSSBytes *uint64 `json:"rssBytes,omitempty"`
// Cumulative number of minor page faults.
// +optional
PageFaults *uint64 `json:"pageFaults,omitempty"`
// Cumulative number of major page faults.
// +optional
MajorPageFaults *uint64 `json:"majorPageFaults,omitempty"`
}
// AcceleratorStats contains stats for accelerators attached to the container.
type AcceleratorStats struct {
// Make of the accelerator (nvidia, amd, google etc.)
Make string `json:"make"`
// Model of the accelerator (tesla-p100, tesla-k80 etc.)
Model string `json:"model"`
// ID of the accelerator.
ID string `json:"id"`
// Total accelerator memory.
// unit: bytes
MemoryTotal uint64 `json:"memoryTotal"`
// Total accelerator memory allocated.
// unit: bytes
MemoryUsed uint64 `json:"memoryUsed"`
// Percent of time over the past sample period (10s) during which
// the accelerator was actively processing.
DutyCycle uint64 `json:"dutyCycle"`
}
// VolumeStats contains data about Volume filesystem usage.
type VolumeStats struct {
// Embedded FsStats
FsStats
// Name is the name given to the Volume
// +optional
Name string `json:"name,omitempty"`
// Reference to the PVC, if one exists
// +optional
PVCRef *PVCReference `json:"pvcRef,omitempty"`
}
// PVCReference contains enough information to describe the referenced PVC.
type PVCReference struct {
Name string `json:"name"`
Namespace string `json:"namespace"`
}
// FsStats contains data about filesystem usage.
type FsStats struct {
// The time at which these stats were updated.
Time metav1.Time `json:"time"`
// AvailableBytes represents the storage space available (bytes) for the filesystem.
// +optional
AvailableBytes *uint64 `json:"availableBytes,omitempty"`
// CapacityBytes represents the total capacity (bytes) of the filesystems underlying storage.
// +optional
CapacityBytes *uint64 `json:"capacityBytes,omitempty"`
// UsedBytes represents the bytes used for a specific task on the filesystem.
// This may differ from the total bytes used on the filesystem and may not equal CapacityBytes - AvailableBytes.
// e.g. For ContainerStats.Rootfs this is the bytes used by the container rootfs on the filesystem.
// +optional
UsedBytes *uint64 `json:"usedBytes,omitempty"`
// InodesFree represents the free inodes in the filesystem.
// +optional
InodesFree *uint64 `json:"inodesFree,omitempty"`
// Inodes represents the total inodes in the filesystem.
// +optional
Inodes *uint64 `json:"inodes,omitempty"`
// InodesUsed represents the inodes used by the filesystem
// This may not equal Inodes - InodesFree because this filesystem may share inodes with other "filesystems"
// e.g. For ContainerStats.Rootfs, this is the inodes used only by that container, and does not count inodes used by other containers.
InodesUsed *uint64 `json:"inodesUsed,omitempty"`
}
// UserDefinedMetricType defines how the metric should be interpreted by the user.
type UserDefinedMetricType string
const (
// MetricGauge is an instantaneous value. May increase or decrease.
MetricGauge UserDefinedMetricType = "gauge"
// MetricCumulative is a counter-like value that is only expected to increase.
MetricCumulative UserDefinedMetricType = "cumulative"
// MetricDelta is a rate over a time period.
MetricDelta UserDefinedMetricType = "delta"
)
// UserDefinedMetricDescriptor contains metadata that describes a user defined metric.
type UserDefinedMetricDescriptor struct {
// The name of the metric.
Name string `json:"name"`
// Type of the metric.
Type UserDefinedMetricType `json:"type"`
// Display Units for the stats.
Units string `json:"units"`
// Metadata labels associated with this metric.
// +optional
Labels map[string]string `json:"labels,omitempty"`
}
// UserDefinedMetric represents a metric defined and generated by users.
type UserDefinedMetric struct {
UserDefinedMetricDescriptor `json:",inline"`
// The time at which these stats were updated.
Time metav1.Time `json:"time"`
// Value of the metric. Float64s have 53 bit precision.
// We do not foresee any metrics exceeding that value.
Value float64 `json:"value"`
}

View File

@ -1,26 +0,0 @@
/*
Copyright 2017 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 apis
const (
// AnnotationProvidedIPAddr is a node IP annotation set by the "external" cloud provider.
// When kubelet is started with the "external" cloud provider, then
// it sets this annotation on the node to denote an ip address set from the
// cmd line flag (--node-ip). This ip is verified with the cloudprovider as valid by
// the cloud-controller-manager
AnnotationProvidedIPAddr = "alpha.kubernetes.io/provided-node-ip"
)

View File

@ -1,43 +0,0 @@
// +build windows
/*
Copyright 2018 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 apis
import (
utilfeature "k8s.io/apiserver/pkg/util/feature"
"k8s.io/kubernetes/pkg/features"
)
const (
// HypervIsolationAnnotationKey is used to run windows containers with hyperv isolation.
// Refer https://aka.ms/hyperv-container.
HypervIsolationAnnotationKey = "experimental.windows.kubernetes.io/isolation-type"
// HypervIsolationValue is used to run windows containers with hyperv isolation.
// Refer https://aka.ms/hyperv-container.
HypervIsolationValue = "hyperv"
)
// ShouldIsolatedByHyperV returns true if a windows container should be run with hyperv isolation.
func ShouldIsolatedByHyperV(annotations map[string]string) bool {
if !utilfeature.DefaultFeatureGate.Enabled(features.HyperVContainer) {
return false
}
v, ok := annotations[HypervIsolationAnnotationKey]
return ok && v == HypervIsolationValue
}

View File

@ -1,91 +0,0 @@
/*
Copyright 2015 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 apis
import (
"strings"
"k8s.io/api/core/v1"
"k8s.io/apimachinery/pkg/util/sets"
)
const (
// LabelOS is a label to indicate the operating system of the node.
// The OS labels are promoted to GA in 1.14. kubelet applies both beta
// and GA labels to ensure backward compatibility.
// TODO: stop applying the beta OS labels in Kubernetes 1.18.
LabelOS = "beta.kubernetes.io/os"
// LabelArch is a label to indicate the architecture of the node.
// The Arch labels are promoted to GA in 1.14. kubelet applies both beta
// and GA labels to ensure backward compatibility.
// TODO: stop applying the beta Arch labels in Kubernetes 1.18.
LabelArch = "beta.kubernetes.io/arch"
)
var kubeletLabels = sets.NewString(
v1.LabelHostname,
v1.LabelZoneFailureDomainStable,
v1.LabelZoneRegionStable,
v1.LabelZoneFailureDomain,
v1.LabelZoneRegion,
v1.LabelInstanceType,
v1.LabelInstanceTypeStable,
v1.LabelOSStable,
v1.LabelArchStable,
LabelOS,
LabelArch,
)
var kubeletLabelNamespaces = sets.NewString(
v1.LabelNamespaceSuffixKubelet,
v1.LabelNamespaceSuffixNode,
)
// KubeletLabels returns the list of label keys kubelets are allowed to set on their own Node objects
func KubeletLabels() []string {
return kubeletLabels.List()
}
// KubeletLabelNamespaces returns the list of label key namespaces kubelets are allowed to set on their own Node objects
func KubeletLabelNamespaces() []string {
return kubeletLabelNamespaces.List()
}
// IsKubeletLabel returns true if the label key is one that kubelets are allowed to set on their own Node object.
// This checks if the key is in the KubeletLabels() list, or has a namespace in the KubeletLabelNamespaces() list.
func IsKubeletLabel(key string) bool {
if kubeletLabels.Has(key) {
return true
}
namespace := getLabelNamespace(key)
for allowedNamespace := range kubeletLabelNamespaces {
if namespace == allowedNamespace || strings.HasSuffix(namespace, "."+allowedNamespace) {
return true
}
}
return false
}
func getLabelNamespace(key string) string {
if parts := strings.SplitN(key, "/", 2); len(parts) == 2 {
return parts[0]
}
return ""
}

View File

@ -1,85 +0,0 @@
load("@io_bazel_rules_go//go:def.bzl", "go_library", "go_test")
go_library(
name = "go_default_library",
srcs = [
"cache.go",
"container_gc.go",
"container_reference_manager.go",
"helpers.go",
"os.go",
"ref.go",
"resize.go",
"runtime.go",
"runtime_cache.go",
"runtime_cache_fake.go",
"sync_result.go",
],
importpath = "k8s.io/kubernetes/pkg/kubelet/container",
visibility = ["//visibility:public"],
deps = [
"//pkg/api/legacyscheme:go_default_library",
"//pkg/api/v1/pod:go_default_library",
"//pkg/features:go_default_library",
"//pkg/kubelet/util/format:go_default_library",
"//pkg/util/hash:go_default_library",
"//pkg/volume:go_default_library",
"//staging/src/k8s.io/api/core/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/runtime:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/types:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/errors:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/runtime:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/util/sets:go_default_library",
"//staging/src/k8s.io/apiserver/pkg/util/feature:go_default_library",
"//staging/src/k8s.io/client-go/tools/record:go_default_library",
"//staging/src/k8s.io/client-go/tools/reference:go_default_library",
"//staging/src/k8s.io/client-go/tools/remotecommand:go_default_library",
"//staging/src/k8s.io/client-go/util/flowcontrol:go_default_library",
"//staging/src/k8s.io/cri-api/pkg/apis/runtime/v1alpha2:go_default_library",
"//third_party/forked/golang/expansion:go_default_library",
"//vendor/k8s.io/klog:go_default_library",
"//vendor/k8s.io/utils/net:go_default_library",
],
)
go_test(
name = "go_default_test",
srcs = [
"cache_test.go",
"container_hash_test.go",
"helpers_test.go",
"ref_test.go",
"runtime_cache_test.go",
"sync_result_test.go",
],
embed = [":go_default_library"],
deps = [
"//pkg/apis/core/install:go_default_library",
"//pkg/features:go_default_library",
"//pkg/kubelet/container/testing:go_default_library",
"//staging/src/k8s.io/api/core/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/apis/meta/v1:go_default_library",
"//staging/src/k8s.io/apimachinery/pkg/types:go_default_library",
"//staging/src/k8s.io/apiserver/pkg/util/feature:go_default_library",
"//staging/src/k8s.io/component-base/featuregate/testing:go_default_library",
"//vendor/github.com/google/go-cmp/cmp:go_default_library",
"//vendor/github.com/stretchr/testify/assert:go_default_library",
],
)
filegroup(
name = "package-srcs",
srcs = glob(["**"]),
tags = ["automanaged"],
visibility = ["//visibility:private"],
)
filegroup(
name = "all-srcs",
srcs = [
":package-srcs",
"//pkg/kubelet/container/testing:all-srcs",
],
tags = ["automanaged"],
visibility = ["//visibility:public"],
)

View File

@ -1,199 +0,0 @@
/*
Copyright 2015 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 container
import (
"sync"
"time"
"k8s.io/apimachinery/pkg/types"
)
// Cache stores the PodStatus for the pods. It represents *all* the visible
// pods/containers in the container runtime. All cache entries are at least as
// new or newer than the global timestamp (set by UpdateTime()), while
// individual entries may be slightly newer than the global timestamp. If a pod
// has no states known by the runtime, Cache returns an empty PodStatus object
// with ID populated.
//
// Cache provides two methods to retrieve the PodStatus: the non-blocking Get()
// and the blocking GetNewerThan() method. The component responsible for
// populating the cache is expected to call Delete() to explicitly free the
// cache entries.
type Cache interface {
Get(types.UID) (*PodStatus, error)
Set(types.UID, *PodStatus, error, time.Time)
// GetNewerThan is a blocking call that only returns the status
// when it is newer than the given time.
GetNewerThan(types.UID, time.Time) (*PodStatus, error)
Delete(types.UID)
UpdateTime(time.Time)
}
type data struct {
// Status of the pod.
status *PodStatus
// Error got when trying to inspect the pod.
err error
// Time when the data was last modified.
modified time.Time
}
type subRecord struct {
time time.Time
ch chan *data
}
// cache implements Cache.
type cache struct {
// Lock which guards all internal data structures.
lock sync.RWMutex
// Map that stores the pod statuses.
pods map[types.UID]*data
// A global timestamp represents how fresh the cached data is. All
// cache content is at the least newer than this timestamp. Note that the
// timestamp is nil after initialization, and will only become non-nil when
// it is ready to serve the cached statuses.
timestamp *time.Time
// Map that stores the subscriber records.
subscribers map[types.UID][]*subRecord
}
// NewCache creates a pod cache.
func NewCache() Cache {
return &cache{pods: map[types.UID]*data{}, subscribers: map[types.UID][]*subRecord{}}
}
// Get returns the PodStatus for the pod; callers are expected not to
// modify the objects returned.
func (c *cache) Get(id types.UID) (*PodStatus, error) {
c.lock.RLock()
defer c.lock.RUnlock()
d := c.get(id)
return d.status, d.err
}
func (c *cache) GetNewerThan(id types.UID, minTime time.Time) (*PodStatus, error) {
ch := c.subscribe(id, minTime)
d := <-ch
return d.status, d.err
}
// Set sets the PodStatus for the pod.
func (c *cache) Set(id types.UID, status *PodStatus, err error, timestamp time.Time) {
c.lock.Lock()
defer c.lock.Unlock()
defer c.notify(id, timestamp)
c.pods[id] = &data{status: status, err: err, modified: timestamp}
}
// Delete removes the entry of the pod.
func (c *cache) Delete(id types.UID) {
c.lock.Lock()
defer c.lock.Unlock()
delete(c.pods, id)
}
// UpdateTime modifies the global timestamp of the cache and notify
// subscribers if needed.
func (c *cache) UpdateTime(timestamp time.Time) {
c.lock.Lock()
defer c.lock.Unlock()
c.timestamp = &timestamp
// Notify all the subscribers if the condition is met.
for id := range c.subscribers {
c.notify(id, *c.timestamp)
}
}
func makeDefaultData(id types.UID) *data {
return &data{status: &PodStatus{ID: id}, err: nil}
}
func (c *cache) get(id types.UID) *data {
d, ok := c.pods[id]
if !ok {
// Cache should store *all* pod/container information known by the
// container runtime. A cache miss indicates that there are no states
// regarding the pod last time we queried the container runtime.
// What this *really* means is that there are no visible pod/containers
// associated with this pod. Simply return an default (mostly empty)
// PodStatus to reflect this.
return makeDefaultData(id)
}
return d
}
// getIfNewerThan returns the data it is newer than the given time.
// Otherwise, it returns nil. The caller should acquire the lock.
func (c *cache) getIfNewerThan(id types.UID, minTime time.Time) *data {
d, ok := c.pods[id]
globalTimestampIsNewer := (c.timestamp != nil && c.timestamp.After(minTime))
if !ok && globalTimestampIsNewer {
// Status is not cached, but the global timestamp is newer than
// minTime, return the default status.
return makeDefaultData(id)
}
if ok && (d.modified.After(minTime) || globalTimestampIsNewer) {
// Status is cached, return status if either of the following is true.
// * status was modified after minTime
// * the global timestamp of the cache is newer than minTime.
return d
}
// The pod status is not ready.
return nil
}
// notify sends notifications for pod with the given id, if the requirements
// are met. Note that the caller should acquire the lock.
func (c *cache) notify(id types.UID, timestamp time.Time) {
list, ok := c.subscribers[id]
if !ok {
// No one to notify.
return
}
newList := []*subRecord{}
for i, r := range list {
if timestamp.Before(r.time) {
// Doesn't meet the time requirement; keep the record.
newList = append(newList, list[i])
continue
}
r.ch <- c.get(id)
close(r.ch)
}
if len(newList) == 0 {
delete(c.subscribers, id)
} else {
c.subscribers[id] = newList
}
}
func (c *cache) subscribe(id types.UID, timestamp time.Time) chan *data {
ch := make(chan *data, 1)
c.lock.Lock()
defer c.lock.Unlock()
d := c.getIfNewerThan(id, timestamp)
if d != nil {
// If the cache entry is ready, send the data and return immediately.
ch <- d
return ch
}
// Add the subscription record.
c.subscribers[id] = append(c.subscribers[id], &subRecord{time: timestamp, ch: ch})
return ch
}

View File

@ -1,87 +0,0 @@
/*
Copyright 2014 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 container
import (
"fmt"
"time"
"k8s.io/klog"
)
// Specified a policy for garbage collecting containers.
type ContainerGCPolicy struct {
// Minimum age at which a container can be garbage collected, zero for no limit.
MinAge time.Duration
// Max number of dead containers any single pod (UID, container name) pair is
// allowed to have, less than zero for no limit.
MaxPerPodContainer int
// Max number of total dead containers, less than zero for no limit.
MaxContainers int
}
// Manages garbage collection of dead containers.
//
// Implementation is thread-compatible.
type ContainerGC interface {
// Garbage collect containers.
GarbageCollect() error
// Deletes all unused containers, including containers belonging to pods that are terminated but not deleted
DeleteAllUnusedContainers() error
}
// SourcesReadyProvider knows how to determine if configuration sources are ready
type SourcesReadyProvider interface {
// AllReady returns true if the currently configured sources have all been seen.
AllReady() bool
}
// TODO(vmarmol): Preferentially remove pod infra containers.
type realContainerGC struct {
// Container runtime
runtime Runtime
// Policy for garbage collection.
policy ContainerGCPolicy
// sourcesReadyProvider provides the readiness of kubelet configuration sources.
sourcesReadyProvider SourcesReadyProvider
}
// New ContainerGC instance with the specified policy.
func NewContainerGC(runtime Runtime, policy ContainerGCPolicy, sourcesReadyProvider SourcesReadyProvider) (ContainerGC, error) {
if policy.MinAge < 0 {
return nil, fmt.Errorf("invalid minimum garbage collection age: %v", policy.MinAge)
}
return &realContainerGC{
runtime: runtime,
policy: policy,
sourcesReadyProvider: sourcesReadyProvider,
}, nil
}
func (cgc *realContainerGC) GarbageCollect() error {
return cgc.runtime.GarbageCollect(cgc.policy, cgc.sourcesReadyProvider.AllReady(), false)
}
func (cgc *realContainerGC) DeleteAllUnusedContainers() error {
klog.Infof("attempting to delete unused containers")
return cgc.runtime.GarbageCollect(cgc.policy, cgc.sourcesReadyProvider.AllReady(), true)
}

View File

@ -1,60 +0,0 @@
/*
Copyright 2015 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 container
import (
"sync"
"k8s.io/api/core/v1"
)
// RefManager manages the references for the containers.
// The references are used for reporting events such as creation,
// failure, etc. This manager is thread-safe, no locks are necessary
// for the caller.
type RefManager struct {
sync.RWMutex
containerIDToRef map[ContainerID]*v1.ObjectReference
}
// NewRefManager creates and returns a container reference manager
// with empty contents.
func NewRefManager() *RefManager {
return &RefManager{containerIDToRef: make(map[ContainerID]*v1.ObjectReference)}
}
// SetRef stores a reference to a pod's container, associating it with the given container ID.
func (c *RefManager) SetRef(id ContainerID, ref *v1.ObjectReference) {
c.Lock()
defer c.Unlock()
c.containerIDToRef[id] = ref
}
// ClearRef forgets the given container id and its associated container reference.
func (c *RefManager) ClearRef(id ContainerID) {
c.Lock()
defer c.Unlock()
delete(c.containerIDToRef, id)
}
// GetRef returns the container reference of the given ID, or (nil, false) if none is stored.
func (c *RefManager) GetRef(id ContainerID) (ref *v1.ObjectReference, ok bool) {
c.RLock()
defer c.RUnlock()
ref, ok = c.containerIDToRef[id]
return ref, ok
}

View File

@ -1,350 +0,0 @@
/*
Copyright 2015 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 container
import (
"encoding/json"
"fmt"
"hash/fnv"
"strings"
"k8s.io/klog"
v1 "k8s.io/api/core/v1"
"k8s.io/apimachinery/pkg/runtime"
"k8s.io/apimachinery/pkg/types"
"k8s.io/apimachinery/pkg/util/sets"
"k8s.io/client-go/tools/record"
runtimeapi "k8s.io/cri-api/pkg/apis/runtime/v1alpha2"
podutil "k8s.io/kubernetes/pkg/api/v1/pod"
"k8s.io/kubernetes/pkg/kubelet/util/format"
hashutil "k8s.io/kubernetes/pkg/util/hash"
"k8s.io/kubernetes/third_party/forked/golang/expansion"
utilsnet "k8s.io/utils/net"
)
// HandlerRunner runs a lifecycle handler for a container.
type HandlerRunner interface {
Run(containerID ContainerID, pod *v1.Pod, container *v1.Container, handler *v1.Handler) (string, error)
}
// RuntimeHelper wraps kubelet to make container runtime
// able to get necessary informations like the RunContainerOptions, DNS settings, Host IP.
type RuntimeHelper interface {
GenerateRunContainerOptions(pod *v1.Pod, container *v1.Container, podIP string, podIPs []string) (contOpts *RunContainerOptions, cleanupAction func(), err error)
GetPodDNS(pod *v1.Pod) (dnsConfig *runtimeapi.DNSConfig, err error)
// GetPodCgroupParent returns the CgroupName identifier, and its literal cgroupfs form on the host
// of a pod.
GetPodCgroupParent(pod *v1.Pod) string
GetPodDir(podUID types.UID) string
GeneratePodHostNameAndDomain(pod *v1.Pod) (hostname string, hostDomain string, err error)
// GetExtraSupplementalGroupsForPod returns a list of the extra
// supplemental groups for the Pod. These extra supplemental groups come
// from annotations on persistent volumes that the pod depends on.
GetExtraSupplementalGroupsForPod(pod *v1.Pod) []int64
}
// ShouldContainerBeRestarted checks whether a container needs to be restarted.
// TODO(yifan): Think about how to refactor this.
func ShouldContainerBeRestarted(container *v1.Container, pod *v1.Pod, podStatus *PodStatus) bool {
// Once a pod has been marked deleted, it should not be restarted
if pod.DeletionTimestamp != nil {
return false
}
// Get latest container status.
status := podStatus.FindContainerStatusByName(container.Name)
// If the container was never started before, we should start it.
// NOTE(random-liu): If all historical containers were GC'd, we'll also return true here.
if status == nil {
return true
}
// Check whether container is running
if status.State == ContainerStateRunning {
return false
}
// Always restart container in the unknown, or in the created state.
if status.State == ContainerStateUnknown || status.State == ContainerStateCreated {
return true
}
// Check RestartPolicy for dead container
if pod.Spec.RestartPolicy == v1.RestartPolicyNever {
klog.V(4).Infof("Already ran container %q of pod %q, do nothing", container.Name, format.Pod(pod))
return false
}
if pod.Spec.RestartPolicy == v1.RestartPolicyOnFailure {
// Check the exit code.
if status.ExitCode == 0 {
klog.V(4).Infof("Already successfully ran container %q of pod %q, do nothing", container.Name, format.Pod(pod))
return false
}
}
return true
}
// HashContainer returns the hash of the container. It is used to compare
// the running container with its desired spec.
// Note: remember to update hashValues in container_hash_test.go as well.
func HashContainer(container *v1.Container) uint64 {
hash := fnv.New32a()
// Omit nil or empty field when calculating hash value
// Please see https://github.com/kubernetes/kubernetes/issues/53644
containerJson, _ := json.Marshal(container)
hashutil.DeepHashObject(hash, containerJson)
return uint64(hash.Sum32())
}
// EnvVarsToMap constructs a map of environment name to value from a slice
// of env vars.
func EnvVarsToMap(envs []EnvVar) map[string]string {
result := map[string]string{}
for _, env := range envs {
result[env.Name] = env.Value
}
return result
}
// V1EnvVarsToMap constructs a map of environment name to value from a slice
// of env vars.
func V1EnvVarsToMap(envs []v1.EnvVar) map[string]string {
result := map[string]string{}
for _, env := range envs {
result[env.Name] = env.Value
}
return result
}
// ExpandContainerCommandOnlyStatic substitutes only static environment variable values from the
// container environment definitions. This does *not* include valueFrom substitutions.
// TODO: callers should use ExpandContainerCommandAndArgs with a fully resolved list of environment.
func ExpandContainerCommandOnlyStatic(containerCommand []string, envs []v1.EnvVar) (command []string) {
mapping := expansion.MappingFuncFor(V1EnvVarsToMap(envs))
if len(containerCommand) != 0 {
for _, cmd := range containerCommand {
command = append(command, expansion.Expand(cmd, mapping))
}
}
return command
}
func ExpandContainerVolumeMounts(mount v1.VolumeMount, envs []EnvVar) (string, error) {
envmap := EnvVarsToMap(envs)
missingKeys := sets.NewString()
expanded := expansion.Expand(mount.SubPathExpr, func(key string) string {
value, ok := envmap[key]
if !ok || len(value) == 0 {
missingKeys.Insert(key)
}
return value
})
if len(missingKeys) > 0 {
return "", fmt.Errorf("missing value for %s", strings.Join(missingKeys.List(), ", "))
}
return expanded, nil
}
func ExpandContainerCommandAndArgs(container *v1.Container, envs []EnvVar) (command []string, args []string) {
mapping := expansion.MappingFuncFor(EnvVarsToMap(envs))
if len(container.Command) != 0 {
for _, cmd := range container.Command {
command = append(command, expansion.Expand(cmd, mapping))
}
}
if len(container.Args) != 0 {
for _, arg := range container.Args {
args = append(args, expansion.Expand(arg, mapping))
}
}
return command, args
}
// Create an event recorder to record object's event except implicitly required container's, like infra container.
func FilterEventRecorder(recorder record.EventRecorder) record.EventRecorder {
return &innerEventRecorder{
recorder: recorder,
}
}
type innerEventRecorder struct {
recorder record.EventRecorder
}
func (irecorder *innerEventRecorder) shouldRecordEvent(object runtime.Object) (*v1.ObjectReference, bool) {
if object == nil {
return nil, false
}
if ref, ok := object.(*v1.ObjectReference); ok {
if !strings.HasPrefix(ref.FieldPath, ImplicitContainerPrefix) {
return ref, true
}
}
return nil, false
}
func (irecorder *innerEventRecorder) Event(object runtime.Object, eventtype, reason, message string) {
if ref, ok := irecorder.shouldRecordEvent(object); ok {
irecorder.recorder.Event(ref, eventtype, reason, message)
}
}
func (irecorder *innerEventRecorder) Eventf(object runtime.Object, eventtype, reason, messageFmt string, args ...interface{}) {
if ref, ok := irecorder.shouldRecordEvent(object); ok {
irecorder.recorder.Eventf(ref, eventtype, reason, messageFmt, args...)
}
}
func (irecorder *innerEventRecorder) AnnotatedEventf(object runtime.Object, annotations map[string]string, eventtype, reason, messageFmt string, args ...interface{}) {
if ref, ok := irecorder.shouldRecordEvent(object); ok {
irecorder.recorder.AnnotatedEventf(ref, annotations, eventtype, reason, messageFmt, args...)
}
}
// Pod must not be nil.
func IsHostNetworkPod(pod *v1.Pod) bool {
return pod.Spec.HostNetwork
}
// TODO(random-liu): Convert PodStatus to running Pod, should be deprecated soon
func ConvertPodStatusToRunningPod(runtimeName string, podStatus *PodStatus) Pod {
runningPod := Pod{
ID: podStatus.ID,
Name: podStatus.Name,
Namespace: podStatus.Namespace,
}
for _, containerStatus := range podStatus.ContainerStatuses {
if containerStatus.State != ContainerStateRunning {
continue
}
container := &Container{
ID: containerStatus.ID,
Name: containerStatus.Name,
Image: containerStatus.Image,
ImageID: containerStatus.ImageID,
Hash: containerStatus.Hash,
State: containerStatus.State,
}
runningPod.Containers = append(runningPod.Containers, container)
}
// Populate sandboxes in kubecontainer.Pod
for _, sandbox := range podStatus.SandboxStatuses {
runningPod.Sandboxes = append(runningPod.Sandboxes, &Container{
ID: ContainerID{Type: runtimeName, ID: sandbox.Id},
State: SandboxToContainerState(sandbox.State),
})
}
return runningPod
}
// SandboxToContainerState converts runtimeapi.PodSandboxState to
// kubecontainer.ContainerState.
// This is only needed because we need to return sandboxes as if they were
// kubecontainer.Containers to avoid substantial changes to PLEG.
// TODO: Remove this once it becomes obsolete.
func SandboxToContainerState(state runtimeapi.PodSandboxState) ContainerState {
switch state {
case runtimeapi.PodSandboxState_SANDBOX_READY:
return ContainerStateRunning
case runtimeapi.PodSandboxState_SANDBOX_NOTREADY:
return ContainerStateExited
}
return ContainerStateUnknown
}
// FormatPod returns a string representing a pod in a human readable format,
// with pod UID as part of the string.
func FormatPod(pod *Pod) string {
// Use underscore as the delimiter because it is not allowed in pod name
// (DNS subdomain format), while allowed in the container name format.
return fmt.Sprintf("%s_%s(%s)", pod.Name, pod.Namespace, pod.ID)
}
// GetContainerSpec gets the container spec by containerName.
func GetContainerSpec(pod *v1.Pod, containerName string) *v1.Container {
var containerSpec *v1.Container
podutil.VisitContainers(&pod.Spec, func(c *v1.Container) bool {
if containerName == c.Name {
containerSpec = c
return false
}
return true
})
return containerSpec
}
// HasPrivilegedContainer returns true if any of the containers in the pod are privileged.
func HasPrivilegedContainer(pod *v1.Pod) bool {
var hasPrivileged bool
podutil.VisitContainers(&pod.Spec, func(c *v1.Container) bool {
if c.SecurityContext != nil && c.SecurityContext.Privileged != nil && *c.SecurityContext.Privileged {
hasPrivileged = true
return false
}
return true
})
return hasPrivileged
}
// MakePortMappings creates internal port mapping from api port mapping.
func MakePortMappings(container *v1.Container) (ports []PortMapping) {
names := make(map[string]struct{})
for _, p := range container.Ports {
pm := PortMapping{
HostPort: int(p.HostPort),
ContainerPort: int(p.ContainerPort),
Protocol: p.Protocol,
HostIP: p.HostIP,
}
// We need to determine the address family this entry applies to. We do this to ensure
// duplicate containerPort / protocol rules work across different address families.
// https://github.com/kubernetes/kubernetes/issues/82373
family := "any"
if p.HostIP != "" {
if utilsnet.IsIPv6String(p.HostIP) {
family = "v6"
} else {
family = "v4"
}
}
// We need to create some default port name if it's not specified, since
// this is necessary for the dockershim CNI driver.
// https://github.com/kubernetes/kubernetes/pull/82374#issuecomment-529496888
if p.Name == "" {
pm.Name = fmt.Sprintf("%s-%s-%s:%d", container.Name, family, p.Protocol, p.ContainerPort)
} else {
pm.Name = fmt.Sprintf("%s-%s", container.Name, p.Name)
}
// Protect against a port name being used more than once in a container.
if _, ok := names[pm.Name]; ok {
klog.Warningf("Port name conflicted, %q is defined more than once", pm.Name)
continue
}
ports = append(ports, pm)
names[pm.Name] = struct{}{}
}
return
}

View File

@ -1,107 +0,0 @@
/*
Copyright 2015 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 container
import (
"io/ioutil"
"os"
"path/filepath"
"time"
)
// OSInterface collects system level operations that need to be mocked out
// during tests.
type OSInterface interface {
MkdirAll(path string, perm os.FileMode) error
Symlink(oldname string, newname string) error
Stat(path string) (os.FileInfo, error)
Remove(path string) error
RemoveAll(path string) error
Create(path string) (*os.File, error)
Chmod(path string, perm os.FileMode) error
Hostname() (name string, err error)
Chtimes(path string, atime time.Time, mtime time.Time) error
Pipe() (r *os.File, w *os.File, err error)
ReadDir(dirname string) ([]os.FileInfo, error)
Glob(pattern string) ([]string, error)
}
// RealOS is used to dispatch the real system level operations.
type RealOS struct{}
// MkdirAll will call os.MkdirAll to create a directory.
func (RealOS) MkdirAll(path string, perm os.FileMode) error {
return os.MkdirAll(path, perm)
}
// Symlink will call os.Symlink to create a symbolic link.
func (RealOS) Symlink(oldname string, newname string) error {
return os.Symlink(oldname, newname)
}
// Stat will call os.Stat to get the FileInfo for a given path
func (RealOS) Stat(path string) (os.FileInfo, error) {
return os.Stat(path)
}
// Remove will call os.Remove to remove the path.
func (RealOS) Remove(path string) error {
return os.Remove(path)
}
// RemoveAll will call os.RemoveAll to remove the path and its children.
func (RealOS) RemoveAll(path string) error {
return os.RemoveAll(path)
}
// Create will call os.Create to create and return a file
// at path.
func (RealOS) Create(path string) (*os.File, error) {
return os.Create(path)
}
// Chmod will change the permissions on the specified path or return
// an error.
func (RealOS) Chmod(path string, perm os.FileMode) error {
return os.Chmod(path, perm)
}
// Hostname will call os.Hostname to return the hostname.
func (RealOS) Hostname() (name string, err error) {
return os.Hostname()
}
// Chtimes will call os.Chtimes to change the atime and mtime of the path
func (RealOS) Chtimes(path string, atime time.Time, mtime time.Time) error {
return os.Chtimes(path, atime, mtime)
}
// Pipe will call os.Pipe to return a connected pair of pipe.
func (RealOS) Pipe() (r *os.File, w *os.File, err error) {
return os.Pipe()
}
// ReadDir will call ioutil.ReadDir to return the files under the directory.
func (RealOS) ReadDir(dirname string) ([]os.FileInfo, error) {
return ioutil.ReadDir(dirname)
}
// Glob will call filepath.Glob to return the names of all files matching
// pattern.
func (RealOS) Glob(pattern string) ([]string, error) {
return filepath.Glob(pattern)
}

View File

@ -1,84 +0,0 @@
/*
Copyright 2015 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 container
import (
"fmt"
"k8s.io/api/core/v1"
utilfeature "k8s.io/apiserver/pkg/util/feature"
ref "k8s.io/client-go/tools/reference"
"k8s.io/kubernetes/pkg/api/legacyscheme"
"k8s.io/kubernetes/pkg/features"
)
var ImplicitContainerPrefix string = "implicitly required container "
// GenerateContainerRef returns an *v1.ObjectReference which references the given container
// within the given pod. Returns an error if the reference can't be constructed or the
// container doesn't actually belong to the pod.
//
// This function will return an error if the provided Pod does not have a selfLink,
// but we expect selfLink to be populated at all call sites for the function.
func GenerateContainerRef(pod *v1.Pod, container *v1.Container) (*v1.ObjectReference, error) {
fieldPath, err := fieldPath(pod, container)
if err != nil {
// TODO: figure out intelligent way to refer to containers that we implicitly
// start (like the pod infra container). This is not a good way, ugh.
fieldPath = ImplicitContainerPrefix + container.Name
}
ref, err := ref.GetPartialReference(legacyscheme.Scheme, pod, fieldPath)
if err != nil {
return nil, err
}
return ref, nil
}
// fieldPath returns a fieldPath locating container within pod.
// Returns an error if the container isn't part of the pod.
func fieldPath(pod *v1.Pod, container *v1.Container) (string, error) {
for i := range pod.Spec.Containers {
here := &pod.Spec.Containers[i]
if here.Name == container.Name {
if here.Name == "" {
return fmt.Sprintf("spec.containers[%d]", i), nil
}
return fmt.Sprintf("spec.containers{%s}", here.Name), nil
}
}
for i := range pod.Spec.InitContainers {
here := &pod.Spec.InitContainers[i]
if here.Name == container.Name {
if here.Name == "" {
return fmt.Sprintf("spec.initContainers[%d]", i), nil
}
return fmt.Sprintf("spec.initContainers{%s}", here.Name), nil
}
}
if utilfeature.DefaultFeatureGate.Enabled(features.EphemeralContainers) {
for i := range pod.Spec.EphemeralContainers {
here := &pod.Spec.EphemeralContainers[i]
if here.Name == container.Name {
if here.Name == "" {
return fmt.Sprintf("spec.ephemeralContainers[%d]", i), nil
}
return fmt.Sprintf("spec.ephemeralContainers{%s}", here.Name), nil
}
}
}
return "", fmt.Errorf("container %q not found in pod %s/%s", container.Name, pod.Namespace, pod.Name)
}

View File

@ -1,42 +0,0 @@
/*
Copyright 2015 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 container
import (
"k8s.io/apimachinery/pkg/util/runtime"
"k8s.io/client-go/tools/remotecommand"
)
// handleResizing spawns a goroutine that processes the resize channel, calling resizeFunc for each
// remotecommand.TerminalSize received from the channel. The resize channel must be closed elsewhere to stop the
// goroutine.
func HandleResizing(resize <-chan remotecommand.TerminalSize, resizeFunc func(size remotecommand.TerminalSize)) {
if resize == nil {
return
}
go func() {
defer runtime.HandleCrash()
for size := range resize {
if size.Height < 1 || size.Width < 1 {
continue
}
resizeFunc(size)
}
}()
}

View File

@ -1,639 +0,0 @@
/*
Copyright 2015 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 container
import (
"context"
"fmt"
"io"
"net/url"
"reflect"
"strings"
"time"
"k8s.io/api/core/v1"
"k8s.io/apimachinery/pkg/types"
"k8s.io/client-go/tools/remotecommand"
"k8s.io/client-go/util/flowcontrol"
runtimeapi "k8s.io/cri-api/pkg/apis/runtime/v1alpha2"
"k8s.io/klog"
"k8s.io/kubernetes/pkg/volume"
)
type Version interface {
// Compare compares two versions of the runtime. On success it returns -1
// if the version is less than the other, 1 if it is greater than the other,
// or 0 if they are equal.
Compare(other string) (int, error)
// String returns a string that represents the version.
String() string
}
// ImageSpec is an internal representation of an image. Currently, it wraps the
// value of a Container's Image field, but in the future it will include more detailed
// information about the different image types.
type ImageSpec struct {
Image string
}
// ImageStats contains statistics about all the images currently available.
type ImageStats struct {
// Total amount of storage consumed by existing images.
TotalStorageBytes uint64
}
// Runtime interface defines the interfaces that should be implemented
// by a container runtime.
// Thread safety is required from implementations of this interface.
type Runtime interface {
// Type returns the type of the container runtime.
Type() string
//SupportsSingleFileMapping returns whether the container runtime supports single file mappings or not.
SupportsSingleFileMapping() bool
// Version returns the version information of the container runtime.
Version() (Version, error)
// APIVersion returns the cached API version information of the container
// runtime. Implementation is expected to update this cache periodically.
// This may be different from the runtime engine's version.
// TODO(random-liu): We should fold this into Version()
APIVersion() (Version, error)
// Status returns the status of the runtime. An error is returned if the Status
// function itself fails, nil otherwise.
Status() (*RuntimeStatus, error)
// GetPods returns a list of containers grouped by pods. The boolean parameter
// specifies whether the runtime returns all containers including those already
// exited and dead containers (used for garbage collection).
GetPods(all bool) ([]*Pod, error)
// GarbageCollect removes dead containers using the specified container gc policy
// If allSourcesReady is not true, it means that kubelet doesn't have the
// complete list of pods from all avialble sources (e.g., apiserver, http,
// file). In this case, garbage collector should refrain itself from aggressive
// behavior such as removing all containers of unrecognized pods (yet).
// If evictNonDeletedPods is set to true, containers and sandboxes belonging to pods
// that are terminated, but not deleted will be evicted. Otherwise, only deleted pods will be GC'd.
// TODO: Revisit this method and make it cleaner.
GarbageCollect(gcPolicy ContainerGCPolicy, allSourcesReady bool, evictNonDeletedPods bool) error
// Syncs the running pod into the desired pod.
SyncPod(pod *v1.Pod, podStatus *PodStatus, pullSecrets []v1.Secret, backOff *flowcontrol.Backoff) PodSyncResult
// KillPod kills all the containers of a pod. Pod may be nil, running pod must not be.
// TODO(random-liu): Return PodSyncResult in KillPod.
// gracePeriodOverride if specified allows the caller to override the pod default grace period.
// only hard kill paths are allowed to specify a gracePeriodOverride in the kubelet in order to not corrupt user data.
// it is useful when doing SIGKILL for hard eviction scenarios, or max grace period during soft eviction scenarios.
KillPod(pod *v1.Pod, runningPod Pod, gracePeriodOverride *int64) error
// GetPodStatus retrieves the status of the pod, including the
// information of all containers in the pod that are visible in Runtime.
GetPodStatus(uid types.UID, name, namespace string) (*PodStatus, error)
// TODO(vmarmol): Unify pod and containerID args.
// GetContainerLogs returns logs of a specific container. By
// default, it returns a snapshot of the container log. Set 'follow' to true to
// stream the log. Set 'follow' to false and specify the number of lines (e.g.
// "100" or "all") to tail the log.
GetContainerLogs(ctx context.Context, pod *v1.Pod, containerID ContainerID, logOptions *v1.PodLogOptions, stdout, stderr io.Writer) (err error)
// Delete a container. If the container is still running, an error is returned.
DeleteContainer(containerID ContainerID) error
// ImageService provides methods to image-related methods.
ImageService
// UpdatePodCIDR sends a new podCIDR to the runtime.
// This method just proxies a new runtimeConfig with the updated
// CIDR value down to the runtime shim.
UpdatePodCIDR(podCIDR string) error
}
// StreamingRuntime is the interface implemented by runtimes that handle the serving of the
// streaming calls (exec/attach/port-forward) themselves. In this case, Kubelet should redirect to
// the runtime server.
type StreamingRuntime interface {
GetExec(id ContainerID, cmd []string, stdin, stdout, stderr, tty bool) (*url.URL, error)
GetAttach(id ContainerID, stdin, stdout, stderr, tty bool) (*url.URL, error)
GetPortForward(podName, podNamespace string, podUID types.UID, ports []int32) (*url.URL, error)
}
type ImageService interface {
// PullImage pulls an image from the network to local storage using the supplied
// secrets if necessary. It returns a reference (digest or ID) to the pulled image.
PullImage(image ImageSpec, pullSecrets []v1.Secret, podSandboxConfig *runtimeapi.PodSandboxConfig) (string, error)
// GetImageRef gets the reference (digest or ID) of the image which has already been in
// the local storage. It returns ("", nil) if the image isn't in the local storage.
GetImageRef(image ImageSpec) (string, error)
// Gets all images currently on the machine.
ListImages() ([]Image, error)
// Removes the specified image.
RemoveImage(image ImageSpec) error
// Returns Image statistics.
ImageStats() (*ImageStats, error)
}
type ContainerAttacher interface {
AttachContainer(id ContainerID, stdin io.Reader, stdout, stderr io.WriteCloser, tty bool, resize <-chan remotecommand.TerminalSize) (err error)
}
type ContainerCommandRunner interface {
// RunInContainer synchronously executes the command in the container, and returns the output.
// If the command completes with a non-0 exit code, a k8s.io/utils/exec.ExitError will be returned.
RunInContainer(id ContainerID, cmd []string, timeout time.Duration) ([]byte, error)
}
// Pod is a group of containers.
type Pod struct {
// The ID of the pod, which can be used to retrieve a particular pod
// from the pod list returned by GetPods().
ID types.UID
// The name and namespace of the pod, which is readable by human.
Name string
Namespace string
// List of containers that belongs to this pod. It may contain only
// running containers, or mixed with dead ones (when GetPods(true)).
Containers []*Container
// List of sandboxes associated with this pod. The sandboxes are converted
// to Container temporariliy to avoid substantial changes to other
// components. This is only populated by kuberuntime.
// TODO: use the runtimeApi.PodSandbox type directly.
Sandboxes []*Container
}
// PodPair contains both runtime#Pod and api#Pod
type PodPair struct {
// APIPod is the v1.Pod
APIPod *v1.Pod
// RunningPod is the pod defined in pkg/kubelet/container/runtime#Pod
RunningPod *Pod
}
// ContainerID is a type that identifies a container.
type ContainerID struct {
// The type of the container runtime. e.g. 'docker'.
Type string
// The identification of the container, this is comsumable by
// the underlying container runtime. (Note that the container
// runtime interface still takes the whole struct as input).
ID string
}
func BuildContainerID(typ, ID string) ContainerID {
return ContainerID{Type: typ, ID: ID}
}
// Convenience method for creating a ContainerID from an ID string.
func ParseContainerID(containerID string) ContainerID {
var id ContainerID
if err := id.ParseString(containerID); err != nil {
klog.Error(err)
}
return id
}
func (c *ContainerID) ParseString(data string) error {
// Trim the quotes and split the type and ID.
parts := strings.Split(strings.Trim(data, "\""), "://")
if len(parts) != 2 {
return fmt.Errorf("invalid container ID: %q", data)
}
c.Type, c.ID = parts[0], parts[1]
return nil
}
func (c *ContainerID) String() string {
return fmt.Sprintf("%s://%s", c.Type, c.ID)
}
func (c *ContainerID) IsEmpty() bool {
return *c == ContainerID{}
}
func (c *ContainerID) MarshalJSON() ([]byte, error) {
return []byte(fmt.Sprintf("%q", c.String())), nil
}
func (c *ContainerID) UnmarshalJSON(data []byte) error {
return c.ParseString(string(data))
}
// DockerID is an ID of docker container. It is a type to make it clear when we're working with docker container Ids
type DockerID string
func (id DockerID) ContainerID() ContainerID {
return ContainerID{
Type: "docker",
ID: string(id),
}
}
type ContainerState string
const (
ContainerStateCreated ContainerState = "created"
ContainerStateRunning ContainerState = "running"
ContainerStateExited ContainerState = "exited"
// This unknown encompasses all the states that we currently don't care.
ContainerStateUnknown ContainerState = "unknown"
)
// Container provides the runtime information for a container, such as ID, hash,
// state of the container.
type Container struct {
// The ID of the container, used by the container runtime to identify
// a container.
ID ContainerID
// The name of the container, which should be the same as specified by
// v1.Container.
Name string
// The image name of the container, this also includes the tag of the image,
// the expected form is "NAME:TAG".
Image string
// The id of the image used by the container.
ImageID string
// Hash of the container, used for comparison. Optional for containers
// not managed by kubelet.
Hash uint64
// State is the state of the container.
State ContainerState
}
// PodStatus represents the status of the pod and its containers.
// v1.PodStatus can be derived from examining PodStatus and v1.Pod.
type PodStatus struct {
// ID of the pod.
ID types.UID
// Name of the pod.
Name string
// Namespace of the pod.
Namespace string
// All IPs assigned to this pod
IPs []string
// Status of containers in the pod.
ContainerStatuses []*ContainerStatus
// Status of the pod sandbox.
// Only for kuberuntime now, other runtime may keep it nil.
SandboxStatuses []*runtimeapi.PodSandboxStatus
}
// ContainerStatus represents the status of a container.
type ContainerStatus struct {
// ID of the container.
ID ContainerID
// Name of the container.
Name string
// Status of the container.
State ContainerState
// Creation time of the container.
CreatedAt time.Time
// Start time of the container.
StartedAt time.Time
// Finish time of the container.
FinishedAt time.Time
// Exit code of the container.
ExitCode int
// Name of the image, this also includes the tag of the image,
// the expected form is "NAME:TAG".
Image string
// ID of the image.
ImageID string
// Hash of the container, used for comparison.
Hash uint64
// Number of times that the container has been restarted.
RestartCount int
// A string explains why container is in such a status.
Reason string
// Message written by the container before exiting (stored in
// TerminationMessagePath).
Message string
}
// FindContainerStatusByName returns container status in the pod status with the given name.
// When there are multiple containers' statuses with the same name, the first match will be returned.
func (podStatus *PodStatus) FindContainerStatusByName(containerName string) *ContainerStatus {
for _, containerStatus := range podStatus.ContainerStatuses {
if containerStatus.Name == containerName {
return containerStatus
}
}
return nil
}
// Get container status of all the running containers in a pod
func (podStatus *PodStatus) GetRunningContainerStatuses() []*ContainerStatus {
runningContainerStatuses := []*ContainerStatus{}
for _, containerStatus := range podStatus.ContainerStatuses {
if containerStatus.State == ContainerStateRunning {
runningContainerStatuses = append(runningContainerStatuses, containerStatus)
}
}
return runningContainerStatuses
}
// Basic information about a container image.
type Image struct {
// ID of the image.
ID string
// Other names by which this image is known.
RepoTags []string
// Digests by which this image is known.
RepoDigests []string
// The size of the image in bytes.
Size int64
}
type EnvVar struct {
Name string
Value string
}
type Annotation struct {
Name string
Value string
}
type Mount struct {
// Name of the volume mount.
// TODO(yifan): Remove this field, as this is not representing the unique name of the mount,
// but the volume name only.
Name string
// Path of the mount within the container.
ContainerPath string
// Path of the mount on the host.
HostPath string
// Whether the mount is read-only.
ReadOnly bool
// Whether the mount needs SELinux relabeling
SELinuxRelabel bool
// Requested propagation mode
Propagation runtimeapi.MountPropagation
}
type PortMapping struct {
// Name of the port mapping
Name string
// Protocol of the port mapping.
Protocol v1.Protocol
// The port number within the container.
ContainerPort int
// The port number on the host.
HostPort int
// The host IP.
HostIP string
}
type DeviceInfo struct {
// Path on host for mapping
PathOnHost string
// Path in Container to map
PathInContainer string
// Cgroup permissions
Permissions string
}
// RunContainerOptions specify the options which are necessary for running containers
type RunContainerOptions struct {
// The environment variables list.
Envs []EnvVar
// The mounts for the containers.
Mounts []Mount
// The host devices mapped into the containers.
Devices []DeviceInfo
// The port mappings for the containers.
PortMappings []PortMapping
// The annotations for the container
// These annotations are generated by other components (i.e.,
// not users). Currently, only device plugins populate the annotations.
Annotations []Annotation
// If the container has specified the TerminationMessagePath, then
// this directory will be used to create and mount the log file to
// container.TerminationMessagePath
PodContainerDir string
// The type of container rootfs
ReadOnly bool
// hostname for pod containers
Hostname string
// EnableHostUserNamespace sets userns=host when users request host namespaces (pid, ipc, net),
// are using non-namespaced capabilities (mknod, sys_time, sys_module), the pod contains a privileged container,
// or using host path volumes.
// This should only be enabled when the container runtime is performing user remapping AND if the
// experimental behavior is desired.
EnableHostUserNamespace bool
}
// VolumeInfo contains information about the volume.
type VolumeInfo struct {
// Mounter is the volume's mounter
Mounter volume.Mounter
// BlockVolumeMapper is the Block volume's mapper
BlockVolumeMapper volume.BlockVolumeMapper
// SELinuxLabeled indicates whether this volume has had the
// pod's SELinux label applied to it or not
SELinuxLabeled bool
// Whether the volume permission is set to read-only or not
// This value is passed from volume.spec
ReadOnly bool
// Inner volume spec name, which is the PV name if used, otherwise
// it is the same as the outer volume spec name.
InnerVolumeSpecName string
}
type VolumeMap map[string]VolumeInfo
// RuntimeConditionType is the types of required runtime conditions.
type RuntimeConditionType string
const (
// RuntimeReady means the runtime is up and ready to accept basic containers.
RuntimeReady RuntimeConditionType = "RuntimeReady"
// NetworkReady means the runtime network is up and ready to accept containers which require network.
NetworkReady RuntimeConditionType = "NetworkReady"
)
// RuntimeStatus contains the status of the runtime.
type RuntimeStatus struct {
// Conditions is an array of current observed runtime conditions.
Conditions []RuntimeCondition
}
// GetRuntimeCondition gets a specified runtime condition from the runtime status.
func (r *RuntimeStatus) GetRuntimeCondition(t RuntimeConditionType) *RuntimeCondition {
for i := range r.Conditions {
c := &r.Conditions[i]
if c.Type == t {
return c
}
}
return nil
}
// String formats the runtime status into human readable string.
func (s *RuntimeStatus) String() string {
var ss []string
for _, c := range s.Conditions {
ss = append(ss, c.String())
}
return fmt.Sprintf("Runtime Conditions: %s", strings.Join(ss, ", "))
}
// RuntimeCondition contains condition information for the runtime.
type RuntimeCondition struct {
// Type of runtime condition.
Type RuntimeConditionType
// Status of the condition, one of true/false.
Status bool
// Reason is brief reason for the condition's last transition.
Reason string
// Message is human readable message indicating details about last transition.
Message string
}
// String formats the runtime condition into human readable string.
func (c *RuntimeCondition) String() string {
return fmt.Sprintf("%s=%t reason:%s message:%s", c.Type, c.Status, c.Reason, c.Message)
}
type Pods []*Pod
// FindPodByID finds and returns a pod in the pod list by UID. It will return an empty pod
// if not found.
func (p Pods) FindPodByID(podUID types.UID) Pod {
for i := range p {
if p[i].ID == podUID {
return *p[i]
}
}
return Pod{}
}
// FindPodByFullName finds and returns a pod in the pod list by the full name.
// It will return an empty pod if not found.
func (p Pods) FindPodByFullName(podFullName string) Pod {
for i := range p {
if BuildPodFullName(p[i].Name, p[i].Namespace) == podFullName {
return *p[i]
}
}
return Pod{}
}
// FindPod combines FindPodByID and FindPodByFullName, it finds and returns a pod in the
// pod list either by the full name or the pod ID. It will return an empty pod
// if not found.
func (p Pods) FindPod(podFullName string, podUID types.UID) Pod {
if len(podFullName) > 0 {
return p.FindPodByFullName(podFullName)
}
return p.FindPodByID(podUID)
}
// FindContainerByName returns a container in the pod with the given name.
// When there are multiple containers with the same name, the first match will
// be returned.
func (p *Pod) FindContainerByName(containerName string) *Container {
for _, c := range p.Containers {
if c.Name == containerName {
return c
}
}
return nil
}
func (p *Pod) FindContainerByID(id ContainerID) *Container {
for _, c := range p.Containers {
if c.ID == id {
return c
}
}
return nil
}
func (p *Pod) FindSandboxByID(id ContainerID) *Container {
for _, c := range p.Sandboxes {
if c.ID == id {
return c
}
}
return nil
}
// ToAPIPod converts Pod to v1.Pod. Note that if a field in v1.Pod has no
// corresponding field in Pod, the field would not be populated.
func (p *Pod) ToAPIPod() *v1.Pod {
var pod v1.Pod
pod.UID = p.ID
pod.Name = p.Name
pod.Namespace = p.Namespace
for _, c := range p.Containers {
var container v1.Container
container.Name = c.Name
container.Image = c.Image
pod.Spec.Containers = append(pod.Spec.Containers, container)
}
return &pod
}
// IsEmpty returns true if the pod is empty.
func (p *Pod) IsEmpty() bool {
return reflect.DeepEqual(p, &Pod{})
}
// GetPodFullName returns a name that uniquely identifies a pod.
func GetPodFullName(pod *v1.Pod) string {
// Use underscore as the delimiter because it is not allowed in pod name
// (DNS subdomain format), while allowed in the container name format.
return pod.Name + "_" + pod.Namespace
}
// Build the pod full name from pod name and namespace.
func BuildPodFullName(name, namespace string) string {
return name + "_" + namespace
}
// Parse the pod full name.
func ParsePodFullName(podFullName string) (string, string, error) {
parts := strings.Split(podFullName, "_")
if len(parts) != 2 || parts[0] == "" || parts[1] == "" {
return "", "", fmt.Errorf("failed to parse the pod full name %q", podFullName)
}
return parts[0], parts[1], nil
}
// Option is a functional option type for Runtime, useful for
// completely optional settings.
type Option func(Runtime)
// Sort the container statuses by creation time.
type SortContainerStatusesByCreationTime []*ContainerStatus
func (s SortContainerStatusesByCreationTime) Len() int { return len(s) }
func (s SortContainerStatusesByCreationTime) Swap(i, j int) { s[i], s[j] = s[j], s[i] }
func (s SortContainerStatusesByCreationTime) Less(i, j int) bool {
return s[i].CreatedAt.Before(s[j].CreatedAt)
}
const (
// MaxPodTerminationMessageLogLength is the maximum bytes any one pod may have written
// as termination message output across all containers. Containers will be evenly truncated
// until output is below this limit.
MaxPodTerminationMessageLogLength = 1024 * 12
// MaxContainerTerminationMessageLength is the upper bound any one container may write to
// its termination message path. Contents above this length will be truncated.
MaxContainerTerminationMessageLength = 1024 * 4
// MaxContainerTerminationMessageLogLength is the maximum bytes any one container will
// have written to its termination message when the message is read from the logs.
MaxContainerTerminationMessageLogLength = 1024 * 2
// MaxContainerTerminationMessageLogLines is the maximum number of previous lines of
// log output that the termination message can contain.
MaxContainerTerminationMessageLogLines = 80
)

View File

@ -1,96 +0,0 @@
/*
Copyright 2015 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 container
import (
"sync"
"time"
)
var (
// TODO(yifan): Maybe set the them as parameters for NewCache().
defaultCachePeriod = time.Second * 2
)
type RuntimeCache interface {
GetPods() ([]*Pod, error)
ForceUpdateIfOlder(time.Time) error
}
type podsGetter interface {
GetPods(bool) ([]*Pod, error)
}
// NewRuntimeCache creates a container runtime cache.
func NewRuntimeCache(getter podsGetter) (RuntimeCache, error) {
return &runtimeCache{
getter: getter,
}, nil
}
// runtimeCache caches a list of pods. It records a timestamp (cacheTime) right
// before updating the pods, so the timestamp is at most as new as the pods
// (and can be slightly older). The timestamp always moves forward. Callers are
// expected not to modify the pods returned from GetPods.
type runtimeCache struct {
sync.Mutex
// The underlying container runtime used to update the cache.
getter podsGetter
// Last time when cache was updated.
cacheTime time.Time
// The content of the cache.
pods []*Pod
}
// GetPods returns the cached pods if they are not outdated; otherwise, it
// retrieves the latest pods and return them.
func (r *runtimeCache) GetPods() ([]*Pod, error) {
r.Lock()
defer r.Unlock()
if time.Since(r.cacheTime) > defaultCachePeriod {
if err := r.updateCache(); err != nil {
return nil, err
}
}
return r.pods, nil
}
func (r *runtimeCache) ForceUpdateIfOlder(minExpectedCacheTime time.Time) error {
r.Lock()
defer r.Unlock()
if r.cacheTime.Before(minExpectedCacheTime) {
return r.updateCache()
}
return nil
}
func (r *runtimeCache) updateCache() error {
pods, timestamp, err := r.getPodsWithTimestamp()
if err != nil {
return err
}
r.pods, r.cacheTime = pods, timestamp
return nil
}
// getPodsWithTimestamp records a timestamp and retrieves pods from the getter.
func (r *runtimeCache) getPodsWithTimestamp() ([]*Pod, time.Time, error) {
// Always record the timestamp before getting the pods to avoid stale pods.
timestamp := time.Now()
pods, err := r.getter.GetPods(false)
return pods, timestamp, err
}

View File

@ -1,45 +0,0 @@
/*
Copyright 2015 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 container
// TestRunTimeCache embeds runtimeCache with some additional methods for testing.
// It must be declared in the container package to have visibility to runtimeCache.
// It cannot be in a "..._test.go" file in order for runtime_cache_test.go to have cross-package visibility to it.
// (cross-package declarations in test files cannot be used from dot imports if this package is vendored)
type TestRuntimeCache struct {
runtimeCache
}
func (r *TestRuntimeCache) UpdateCacheWithLock() error {
r.Lock()
defer r.Unlock()
return r.updateCache()
}
func (r *TestRuntimeCache) GetCachedPods() []*Pod {
r.Lock()
defer r.Unlock()
return r.pods
}
func NewTestRuntimeCache(getter podsGetter) *TestRuntimeCache {
return &TestRuntimeCache{
runtimeCache: runtimeCache{
getter: getter,
},
}
}

View File

@ -1,128 +0,0 @@
/*
Copyright 2015 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 container
import (
"errors"
"fmt"
utilerrors "k8s.io/apimachinery/pkg/util/errors"
)
// TODO(random-liu): We need to better organize runtime errors for introspection.
// Container Terminated and Kubelet is backing off the restart
var ErrCrashLoopBackOff = errors.New("CrashLoopBackOff")
var (
// ErrContainerNotFound returned when a container in the given pod with the
// given container name was not found, amongst those managed by the kubelet.
ErrContainerNotFound = errors.New("no matching container")
)
var (
ErrRunContainer = errors.New("RunContainerError")
ErrKillContainer = errors.New("KillContainerError")
ErrVerifyNonRoot = errors.New("VerifyNonRootError")
ErrRunInitContainer = errors.New("RunInitContainerError")
ErrCreatePodSandbox = errors.New("CreatePodSandboxError")
ErrConfigPodSandbox = errors.New("ConfigPodSandboxError")
ErrKillPodSandbox = errors.New("KillPodSandboxError")
)
var (
ErrSetupNetwork = errors.New("SetupNetworkError")
ErrTeardownNetwork = errors.New("TeardownNetworkError")
)
// SyncAction indicates different kind of actions in SyncPod() and KillPod(). Now there are only actions
// about start/kill container and setup/teardown network.
type SyncAction string
const (
StartContainer SyncAction = "StartContainer"
KillContainer SyncAction = "KillContainer"
SetupNetwork SyncAction = "SetupNetwork"
TeardownNetwork SyncAction = "TeardownNetwork"
InitContainer SyncAction = "InitContainer"
CreatePodSandbox SyncAction = "CreatePodSandbox"
ConfigPodSandbox SyncAction = "ConfigPodSandbox"
KillPodSandbox SyncAction = "KillPodSandbox"
)
// SyncResult is the result of sync action.
type SyncResult struct {
// The associated action of the result
Action SyncAction
// The target of the action, now the target can only be:
// * Container: Target should be container name
// * Network: Target is useless now, we just set it as pod full name now
Target interface{}
// Brief error reason
Error error
// Human readable error reason
Message string
}
// NewSyncResult generates new SyncResult with specific Action and Target
func NewSyncResult(action SyncAction, target interface{}) *SyncResult {
return &SyncResult{Action: action, Target: target}
}
// Fail fails the SyncResult with specific error and message
func (r *SyncResult) Fail(err error, msg string) {
r.Error, r.Message = err, msg
}
// PodSyncResult is the summary result of SyncPod() and KillPod()
type PodSyncResult struct {
// Result of different sync actions
SyncResults []*SyncResult
// Error encountered in SyncPod() and KillPod() that is not already included in SyncResults
SyncError error
}
// AddSyncResult adds multiple SyncResult to current PodSyncResult
func (p *PodSyncResult) AddSyncResult(result ...*SyncResult) {
p.SyncResults = append(p.SyncResults, result...)
}
// AddPodSyncResult merges a PodSyncResult to current one
func (p *PodSyncResult) AddPodSyncResult(result PodSyncResult) {
p.AddSyncResult(result.SyncResults...)
p.SyncError = result.SyncError
}
// Fail fails the PodSyncResult with an error occurred in SyncPod() and KillPod() itself
func (p *PodSyncResult) Fail(err error) {
p.SyncError = err
}
// Error returns an error summarizing all the errors in PodSyncResult
func (p *PodSyncResult) Error() error {
errlist := []error{}
if p.SyncError != nil {
errlist = append(errlist, fmt.Errorf("failed to SyncPod: %v", p.SyncError))
}
for _, result := range p.SyncResults {
if result.Error != nil {
errlist = append(errlist, fmt.Errorf("failed to %q for %q with %v: %q", result.Action, result.Target,
result.Error, result.Message))
}
}
return utilerrors.NewAggregate(errlist)
}

View File

@ -1,26 +0,0 @@
load("@io_bazel_rules_go//go:def.bzl", "go_library")
go_library(
name = "go_default_library",
srcs = ["metrics.go"],
importpath = "k8s.io/kubernetes/pkg/kubelet/dockershim/metrics",
visibility = ["//visibility:public"],
deps = [
"//staging/src/k8s.io/component-base/metrics:go_default_library",
"//staging/src/k8s.io/component-base/metrics/legacyregistry:go_default_library",
],
)
filegroup(
name = "package-srcs",
srcs = glob(["**"]),
tags = ["automanaged"],
visibility = ["//visibility:private"],
)
filegroup(
name = "all-srcs",
srcs = [":package-srcs"],
tags = ["automanaged"],
visibility = ["//visibility:public"],
)

Some files were not shown because too many files have changed in this diff Show More