package util

import "k8s.io/kubernetes/pkg/scheduler/util"

Index ¶

• Constants
• func As[T any](oldObj, newobj interface{}) (T, T, error)
• func ClearNominatedNodeName(ctx context.Context, cs kubernetes.Interface, pods ...*v1.Pod) utilerrors.Aggregate
• func DeletePod(ctx context.Context, cs kubernetes.Interface, pod *v1.Pod) error
• func GetEarliestPodStartTime(victims *extenderv1.Victims) *metav1.Time
• func GetPodFullName(pod *v1.Pod) string
• func GetPodStartTime(pod *v1.Pod) *metav1.Time
• func IsScalarResourceName(name v1.ResourceName) bool
• func MoreImportantPod(pod1, pod2 *v1.Pod) bool
• func PatchPodStatus(ctx context.Context, cs kubernetes.Interface, old *v1.Pod, newStatus *v1.PodStatus) error
• func Retriable(err error) bool

Constants ¶

const (
	// DefaultMilliCPURequest defines default milli cpu request number.
	DefaultMilliCPURequest int64 = 100 // 0.1 core
	// DefaultMemoryRequest defines default memory request size.
	DefaultMemoryRequest int64 = 200 * 1024 * 1024 // 200 MB
)

For each of these resources, a container that doesn't request the resource explicitly will be treated as having requested the amount indicated below, for the purpose of computing priority only. This ensures that when scheduling zero-request pods, such pods will not all be scheduled to the node with the smallest in-use request, and that when scheduling regular pods, such pods will not see zero-request pods as consuming no resources whatsoever. We chose these values to be similar to the resources that we give to cluster addon pods (#10653). But they are pretty arbitrary. As described in #11713, we use request instead of limit to deal with resource requirements.

Functions ¶

func As ¶

func As[T any](oldObj, newobj interface{}) (T, T, error)

As converts two objects to the given type. Both objects must be of the same type. If not, an error is returned. nil objects are allowed and will be converted to nil. For oldObj, cache.DeletedFinalStateUnknown is handled and the object stored in it will be converted instead.

func ClearNominatedNodeName ¶

func ClearNominatedNodeName(ctx context.Context, cs kubernetes.Interface, pods ...*v1.Pod) utilerrors.Aggregate

ClearNominatedNodeName internally submit a patch request to API server to set each pods[*].Status.NominatedNodeName> to "".

func DeletePod ¶

func DeletePod(ctx context.Context, cs kubernetes.Interface, pod *v1.Pod) error

DeletePod deletes the given <pod> from API server

func GetEarliestPodStartTime ¶

func GetEarliestPodStartTime(victims *extenderv1.Victims) *metav1.Time

GetEarliestPodStartTime returns the earliest start time of all pods that have the highest priority among all victims.

func GetPodFullName ¶

func GetPodFullName(pod *v1.Pod) string

GetPodFullName returns a name that uniquely identifies a pod.

func GetPodStartTime ¶

func GetPodStartTime(pod *v1.Pod) *metav1.Time

GetPodStartTime returns start time of the given pod or current timestamp if it hasn't started yet.

func IsScalarResourceName ¶

func IsScalarResourceName(name v1.ResourceName) bool

IsScalarResourceName validates the resource for Extended, Hugepages, Native and AttachableVolume resources

func MoreImportantPod ¶

func MoreImportantPod(pod1, pod2 *v1.Pod) bool

MoreImportantPod return true when priority of the first pod is higher than the second one. If two pods' priorities are equal, compare their StartTime. It takes arguments of the type "interface{}" to be used with SortableList, but expects those arguments to be *v1.Pod.

func PatchPodStatus ¶

func PatchPodStatus(ctx context.Context, cs kubernetes.Interface, old *v1.Pod, newStatus *v1.PodStatus) error

PatchPodStatus calculates the delta bytes change from <old.Status> to <newStatus>, and then submit a request to API server to patch the pod changes.

func Retriable ¶

func Retriable(err error) bool

Retriable defines the retriable errors during a scheduling cycle.

Source Files ¶

pod_resources.go utils.go

Directories ¶