Move container status AllocatedResources behind a separate feature gate

This commit is contained in:
Tim Allclair 2024-10-25 15:14:03 -07:00
parent 3f5d0ee2cf
commit 0f0e27d226
6 changed files with 132 additions and 79 deletions

View File

@ -828,18 +828,29 @@ func dropDisabledPodStatusFields(podStatus, oldPodStatus *api.PodStatus, podSpec
}
if !utilfeature.DefaultFeatureGate.Enabled(features.InPlacePodVerticalScaling) && !inPlacePodVerticalScalingInUse(oldPodSpec) {
// Drop Resize, AllocatedResources, and Resources fields
dropResourcesFields := func(csl []api.ContainerStatus) {
// Drop Resize and Resources fields
dropResourcesField := func(csl []api.ContainerStatus) {
for i := range csl {
csl[i].AllocatedResources = nil
csl[i].Resources = nil
}
}
dropResourcesFields(podStatus.ContainerStatuses)
dropResourcesFields(podStatus.InitContainerStatuses)
dropResourcesFields(podStatus.EphemeralContainerStatuses)
dropResourcesField(podStatus.ContainerStatuses)
dropResourcesField(podStatus.InitContainerStatuses)
dropResourcesField(podStatus.EphemeralContainerStatuses)
podStatus.Resize = ""
}
if !utilfeature.DefaultFeatureGate.Enabled(features.InPlacePodVerticalScaling) ||
!utilfeature.DefaultFeatureGate.Enabled(features.InPlacePodVerticalScalingAllocatedStatus) {
// Drop AllocatedResources field
dropAllocatedResourcesField := func(csl []api.ContainerStatus) {
for i := range csl {
csl[i].AllocatedResources = nil
}
}
dropAllocatedResourcesField(podStatus.ContainerStatuses)
dropAllocatedResourcesField(podStatus.InitContainerStatuses)
dropAllocatedResourcesField(podStatus.EphemeralContainerStatuses)
}
if !utilfeature.DefaultFeatureGate.Enabled(features.DynamicResourceAllocation) && !dynamicResourceAllocationInUse(oldPodSpec) {
podStatus.ResourceClaimStatuses = nil

View File

@ -2635,56 +2635,69 @@ func TestDropInPlacePodVerticalScaling(t *testing.T) {
},
}
for _, enabled := range []bool{true, false} {
for _, oldPodInfo := range podInfo {
for _, newPodInfo := range podInfo {
oldPodHasInPlaceVerticalScaling, oldPod := oldPodInfo.hasInPlaceVerticalScaling, oldPodInfo.pod()
newPodHasInPlaceVerticalScaling, newPod := newPodInfo.hasInPlaceVerticalScaling, newPodInfo.pod()
if newPod == nil {
continue
}
for _, ippvsEnabled := range []bool{true, false} {
t.Run(fmt.Sprintf("InPlacePodVerticalScaling=%t", ippvsEnabled), func(t *testing.T) {
featuregatetesting.SetFeatureGateDuringTest(t, utilfeature.DefaultFeatureGate, features.InPlacePodVerticalScaling, ippvsEnabled)
t.Run(fmt.Sprintf("feature enabled=%v, old pod %v, new pod %v", enabled, oldPodInfo.description, newPodInfo.description), func(t *testing.T) {
featuregatetesting.SetFeatureGateDuringTest(t, utilfeature.DefaultFeatureGate, features.InPlacePodVerticalScaling, enabled)
for _, allocatedStatusEnabled := range []bool{true, false} {
t.Run(fmt.Sprintf("AllocatedStatus=%t", allocatedStatusEnabled), func(t *testing.T) {
featuregatetesting.SetFeatureGateDuringTest(t, utilfeature.DefaultFeatureGate, features.InPlacePodVerticalScalingAllocatedStatus, allocatedStatusEnabled)
var oldPodSpec *api.PodSpec
var oldPodStatus *api.PodStatus
if oldPod != nil {
oldPodSpec = &oldPod.Spec
oldPodStatus = &oldPod.Status
}
dropDisabledFields(&newPod.Spec, nil, oldPodSpec, nil)
dropDisabledPodStatusFields(&newPod.Status, oldPodStatus, &newPod.Spec, oldPodSpec)
for _, oldPodInfo := range podInfo {
for _, newPodInfo := range podInfo {
oldPodHasInPlaceVerticalScaling, oldPod := oldPodInfo.hasInPlaceVerticalScaling, oldPodInfo.pod()
newPodHasInPlaceVerticalScaling, newPod := newPodInfo.hasInPlaceVerticalScaling, newPodInfo.pod()
if newPod == nil {
continue
}
// old pod should never be changed
if !reflect.DeepEqual(oldPod, oldPodInfo.pod()) {
t.Errorf("old pod changed: %v", cmp.Diff(oldPod, oldPodInfo.pod()))
}
t.Run(fmt.Sprintf("old pod %v, new pod %v", oldPodInfo.description, newPodInfo.description), func(t *testing.T) {
var oldPodSpec *api.PodSpec
var oldPodStatus *api.PodStatus
if oldPod != nil {
oldPodSpec = &oldPod.Spec
oldPodStatus = &oldPod.Status
}
dropDisabledFields(&newPod.Spec, nil, oldPodSpec, nil)
dropDisabledPodStatusFields(&newPod.Status, oldPodStatus, &newPod.Spec, oldPodSpec)
switch {
case enabled || oldPodHasInPlaceVerticalScaling:
// new pod shouldn't change if feature enabled or if old pod has ResizePolicy set
if !reflect.DeepEqual(newPod, newPodInfo.pod()) {
t.Errorf("new pod changed: %v", cmp.Diff(newPod, newPodInfo.pod()))
}
case newPodHasInPlaceVerticalScaling:
// new pod should be changed
if reflect.DeepEqual(newPod, newPodInfo.pod()) {
t.Errorf("new pod was not changed")
}
// new pod should not have ResizePolicy
if !reflect.DeepEqual(newPod, podWithoutInPlaceVerticalScaling()) {
t.Errorf("new pod has ResizePolicy: %v", cmp.Diff(newPod, podWithoutInPlaceVerticalScaling()))
}
default:
// new pod should not need to be changed
if !reflect.DeepEqual(newPod, newPodInfo.pod()) {
t.Errorf("new pod changed: %v", cmp.Diff(newPod, newPodInfo.pod()))
// old pod should never be changed
if !reflect.DeepEqual(oldPod, oldPodInfo.pod()) {
t.Errorf("old pod changed: %v", cmp.Diff(oldPod, oldPodInfo.pod()))
}
switch {
case ippvsEnabled || oldPodHasInPlaceVerticalScaling:
// new pod shouldn't change if feature enabled or if old pod has ResizePolicy set
expected := newPodInfo.pod()
if !ippvsEnabled || !allocatedStatusEnabled {
expected.Status.ContainerStatuses[0].AllocatedResources = nil
}
if !reflect.DeepEqual(newPod, expected) {
t.Errorf("new pod changed: %v", cmp.Diff(newPod, expected))
}
case newPodHasInPlaceVerticalScaling:
// new pod should be changed
if reflect.DeepEqual(newPod, newPodInfo.pod()) {
t.Errorf("new pod was not changed")
}
// new pod should not have ResizePolicy
if !reflect.DeepEqual(newPod, podWithoutInPlaceVerticalScaling()) {
t.Errorf("new pod has ResizePolicy: %v", cmp.Diff(newPod, podWithoutInPlaceVerticalScaling()))
}
default:
// new pod should not need to be changed
if !reflect.DeepEqual(newPod, newPodInfo.pod()) {
t.Errorf("new pod changed: %v", cmp.Diff(newPod, newPodInfo.pod()))
}
}
})
}
}
})
}
}
})
}
}

View File

@ -264,6 +264,19 @@ const (
// deletion ordering.
HonorPVReclaimPolicy featuregate.Feature = "HonorPVReclaimPolicy"
// owner: @vinaykul,@tallclair
// kep: http://kep.k8s.io/1287
//
// Enables In-Place Pod Vertical Scaling
InPlacePodVerticalScaling featuregate.Feature = "InPlacePodVerticalScaling"
// owner: @tallclair
// kep: http://kep.k8s.io/1287
//
// Enables the AllocatedResources field in container status. This feature requires
// InPlacePodVerticalScaling also be enabled.
InPlacePodVerticalScalingAllocatedStatus featuregate.Feature = "InPlacePodVerticalScalingAllocatedStatus"
// owner: @trierra
//
// Disables the Portworx in-tree driver.
@ -733,12 +746,6 @@ const (
// Initial implementation focused on ReadWriteOncePod volumes.
SELinuxMountReadWriteOncePod featuregate.Feature = "SELinuxMountReadWriteOncePod"
// owner: @vinaykul
// kep: http://kep.k8s.io/1287
//
// Enables In-Place Pod Vertical Scaling
InPlacePodVerticalScaling featuregate.Feature = "InPlacePodVerticalScaling"
// owner: @Sh4d1,@RyanAoh,@rikatz
// kep: http://kep.k8s.io/1860
// LoadBalancerIPMode enables the IPMode field in the LoadBalancerIngress status of a Service

View File

@ -391,6 +391,10 @@ var defaultVersionedKubernetesFeatureGates = map[featuregate.Feature]featuregate
{Version: version.MustParse("1.27"), Default: false, PreRelease: featuregate.Alpha},
},
InPlacePodVerticalScalingAllocatedStatus: {
{Version: version.MustParse("1.32"), Default: false, PreRelease: featuregate.Alpha},
},
InTreePluginPortworxUnregister: {
{Version: version.MustParse("1.23"), Default: false, PreRelease: featuregate.Alpha},
},

View File

@ -2095,18 +2095,14 @@ func (kl *Kubelet) convertToAPIContainerStatuses(pod *v1.Pod, podStatus *kubecon
}
}
}
container := kubecontainer.GetContainerSpec(pod, cName)
// Always set the status to the latest allocated resources, even if it differs from the
// allocation used by the current sync loop.
alloc, found := kl.statusManager.GetContainerResourceAllocation(string(pod.UID), cName)
if found {
status.AllocatedResources = alloc.Requests
} else if !(container.Resources.Requests == nil && container.Resources.Limits == nil) {
// This case is expected for ephemeral containers.
if oldStatusFound {
status.AllocatedResources = oldStatus.AllocatedResources
}
if !found {
// This case is expected for non-resizeable containers.
// Don't set status.Resources in this case.
return nil
}
if oldStatus.Resources == nil {
oldStatus.Resources = &v1.ResourceRequirements{}
@ -2342,6 +2338,12 @@ func (kl *Kubelet) convertToAPIContainerStatuses(pod *v1.Pod, podStatus *kubecon
if status.State.Running != nil {
status.Resources = convertContainerStatusResources(cName, status, cStatus, oldStatuses)
}
if utilfeature.DefaultFeatureGate.Enabled(features.InPlacePodVerticalScalingAllocatedStatus) {
if alloc, found := kl.statusManager.GetContainerResourceAllocation(string(pod.UID), cName); found {
status.AllocatedResources = alloc.Requests
}
}
}
if utilfeature.DefaultFeatureGate.Enabled(features.SupplementalGroupsPolicy) {
status.User = convertContainerStatusUser(cStatus)

View File

@ -4543,6 +4543,7 @@ func TestConvertToAPIContainerStatusesDataRace(t *testing.T) {
func TestConvertToAPIContainerStatusesForResources(t *testing.T) {
featuregatetesting.SetFeatureGateDuringTest(t, utilfeature.DefaultFeatureGate, features.InPlacePodVerticalScaling, true)
nowTime := time.Now()
testContainerName := "ctr0"
testContainerID := kubecontainer.ContainerID{Type: "test", ID: testContainerName}
@ -4843,26 +4844,41 @@ func TestConvertToAPIContainerStatusesForResources(t *testing.T) {
},
},
} {
tPod := testPod.DeepCopy()
tPod.Name = fmt.Sprintf("%s-%d", testPod.Name, idx)
for i := range tPod.Spec.Containers {
if tc.Resources != nil {
tPod.Spec.Containers[i].Resources = tc.Resources[i]
}
kubelet.statusManager.SetPodAllocation(tPod)
if tc.Resources != nil {
tPod.Status.ContainerStatuses[i].AllocatedResources = tc.Resources[i].Requests
testPodStatus.ContainerStatuses[i].Resources = &kubecontainer.ContainerResources{
MemoryLimit: tc.Resources[i].Limits.Memory(),
CPULimit: tc.Resources[i].Limits.Cpu(),
CPURequest: tc.Resources[i].Requests.Cpu(),
t.Run(tdesc, func(t *testing.T) {
tPod := testPod.DeepCopy()
tPod.Name = fmt.Sprintf("%s-%d", testPod.Name, idx)
for i := range tPod.Spec.Containers {
if tc.Resources != nil {
tPod.Spec.Containers[i].Resources = tc.Resources[i]
}
kubelet.statusManager.SetPodAllocation(tPod)
if tc.Resources != nil {
testPodStatus.ContainerStatuses[i].Resources = &kubecontainer.ContainerResources{
MemoryLimit: tc.Resources[i].Limits.Memory(),
CPULimit: tc.Resources[i].Limits.Cpu(),
CPURequest: tc.Resources[i].Requests.Cpu(),
}
}
}
}
t.Logf("TestCase: %q", tdesc)
cStatuses := kubelet.convertToAPIContainerStatuses(tPod, testPodStatus, tc.OldStatus, tPod.Spec.Containers, false, false)
assert.Equal(t, tc.Expected, cStatuses)
for _, enableAllocatedStatus := range []bool{true, false} {
t.Run(fmt.Sprintf("AllocatedStatus=%t", enableAllocatedStatus), func(t *testing.T) {
featuregatetesting.SetFeatureGateDuringTest(t, utilfeature.DefaultFeatureGate, features.InPlacePodVerticalScalingAllocatedStatus, enableAllocatedStatus)
expected := tc.Expected
if !enableAllocatedStatus {
for i, status := range expected {
noAllocated := *status.DeepCopy()
noAllocated.AllocatedResources = nil
expected[i] = noAllocated
}
}
cStatuses := kubelet.convertToAPIContainerStatuses(tPod, testPodStatus, tc.OldStatus, tPod.Spec.Containers, false, false)
assert.Equal(t, expected, cStatuses)
})
}
})
}
}