kubernetes – k8s.io/kubernetes/test/e2e/scheduling Index | Files

package scheduling

import "k8s.io/kubernetes/test/e2e/scheduling"

Index

Functions

func CreateHostPortPods 

func CreateHostPortPods(ctx context.Context, f *framework.Framework, id string, replicas int, expectRunning bool)

CreateHostPortPods creates RC with host port 4321

func CreateNodeSelectorPods 

func CreateNodeSelectorPods(ctx context.Context, f *framework.Framework, id string, replicas int, nodeSelector map[string]string, expectRunning bool) error

CreateNodeSelectorPods creates RC with host port 4321 and defines node selector

func Get2NodesThatCanRunPod 

func Get2NodesThatCanRunPod(ctx context.Context, f *framework.Framework) []string

Get2NodesThatCanRunPod return a 2-node slice where can run pod.

func GetNodeThatCanRunPod 

func GetNodeThatCanRunPod(ctx context.Context, f *framework.Framework) string

GetNodeThatCanRunPod trying to launch a pod without a label to get a node which can launch it

func GetPodsScheduled 

func GetPodsScheduled(workerNodes sets.Set[string], pods *v1.PodList) (scheduledPods, notScheduledPods []v1.Pod)

GetPodsScheduled returns a number of currently scheduled and not scheduled Pods on worker nodes.

func SIGDescribe 

func SIGDescribe(text string, body func()) bool

SIGDescribe annotates the test with the SIG label.

func SetupNVIDIAGPUNode 

func SetupNVIDIAGPUNode(ctx context.Context, f *framework.Framework, setupResourceGatherer bool) *e2edebug.ContainerResourceGatherer

SetupNVIDIAGPUNode install Nvidia Drivers and wait for Nvidia GPUs to be available on nodes

func SpreadRCOrFail 

func SpreadRCOrFail(ctx context.Context, f *framework.Framework, replicaCount int32, zoneNames sets.Set[string], image string, args []string)

SpreadRCOrFail Check that the pods comprising a replication controller get spread evenly across available zones

func SpreadServiceOrFail 

func SpreadServiceOrFail(ctx context.Context, f *framework.Framework, replicaCount int, zoneNames sets.Set[string], image string)

SpreadServiceOrFail check that the pods comprising a service get spread evenly across available zones

func StartJob 

func StartJob(ctx context.Context, f *framework.Framework, completions int32)

StartJob starts a simple CUDA job that requests gpu and the specified number of completions

func VerifyJobNCompletions 

func VerifyJobNCompletions(ctx context.Context, f *framework.Framework, completions int32)

VerifyJobNCompletions verifies that the job has completions number of successful pods

func WaitForSchedulerAfterAction 

func WaitForSchedulerAfterAction(ctx context.Context, f *framework.Framework, action Action, ns, podName string, expectSuccess bool)

WaitForSchedulerAfterAction performs the provided action and then waits for scheduler to act on the given pod.

func WaitForStableCluster 

func WaitForStableCluster(c clientset.Interface, workerNodes sets.Set[string]) int

WaitForStableCluster waits until all existing pods are scheduled and returns their amount.

Types

type Action 

type Action func(ctx context.Context) error

Action is a function to be performed by the system.

type Resource 

type Resource struct {
	MilliCPU int64
	Memory   int64
}

Resource is a collection of compute resource.

Source Files

events.go framework.go limit_range.go nvidia-gpus.go predicates.go preemption.go priorities.go ubernetes_lite.go

Version
v1.29.0-alpha.1
Published
Sep 25, 2023
Platform
js/wasm
Imports
62 packages
Last checked
3 minutes ago –

Tools for package owners.