vendor update for E2E framework

Signed-off-by: Madhu Rajanna <madhupr007@gmail.com>

vendor/k8s.io/kubernetes/pkg/scheduler/algorithm/doc.go

@@ -0,0 +1,19 @@
+/*
+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 algorithm contains a generic Scheduler interface and several
+// implementations.
+package algorithm // import "k8s.io/kubernetes/pkg/scheduler/algorithm"

vendor/k8s.io/kubernetes/pkg/scheduler/algorithm/predicates/csi_volume_predicate.go

@@ -0,0 +1,156 @@
+/*
+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 predicates
+
+import (
+	"fmt"
+
+	"k8s.io/api/core/v1"
+	utilfeature "k8s.io/apiserver/pkg/util/feature"
+	"k8s.io/klog"
+	"k8s.io/kubernetes/pkg/features"
+	schedulernodeinfo "k8s.io/kubernetes/pkg/scheduler/nodeinfo"
+	volumeutil "k8s.io/kubernetes/pkg/volume/util"
+)
+
+// CSIMaxVolumeLimitChecker defines predicate needed for counting CSI volumes
+type CSIMaxVolumeLimitChecker struct {
+	pvInfo  PersistentVolumeInfo
+	pvcInfo PersistentVolumeClaimInfo
+}
+
+// NewCSIMaxVolumeLimitPredicate returns a predicate for counting CSI volumes
+func NewCSIMaxVolumeLimitPredicate(
+	pvInfo PersistentVolumeInfo, pvcInfo PersistentVolumeClaimInfo) FitPredicate {
+	c := &CSIMaxVolumeLimitChecker{
+		pvInfo:  pvInfo,
+		pvcInfo: pvcInfo,
+	}
+	return c.attachableLimitPredicate
+}
+
+func (c *CSIMaxVolumeLimitChecker) attachableLimitPredicate(
+	pod *v1.Pod, meta PredicateMetadata, nodeInfo *schedulernodeinfo.NodeInfo) (bool, []PredicateFailureReason, error) {
+
+	// if feature gate is disable we return
+	if !utilfeature.DefaultFeatureGate.Enabled(features.AttachVolumeLimit) {
+		return true, nil, nil
+	}
+	// If a pod doesn't have any volume attached to it, the predicate will always be true.
+	// Thus we make a fast path for it, to avoid unnecessary computations in this case.
+	if len(pod.Spec.Volumes) == 0 {
+		return true, nil, nil
+	}
+
+	nodeVolumeLimits := nodeInfo.VolumeLimits()
+
+	// if node does not have volume limits this predicate should exit
+	if len(nodeVolumeLimits) == 0 {
+		return true, nil, nil
+	}
+
+	// a map of unique volume name/csi volume handle and volume limit key
+	newVolumes := make(map[string]string)
+	if err := c.filterAttachableVolumes(pod.Spec.Volumes, pod.Namespace, newVolumes); err != nil {
+		return false, nil, err
+	}
+
+	if len(newVolumes) == 0 {
+		return true, nil, nil
+	}
+
+	// a map of unique volume name/csi volume handle and volume limit key
+	attachedVolumes := make(map[string]string)
+	for _, existingPod := range nodeInfo.Pods() {
+		if err := c.filterAttachableVolumes(existingPod.Spec.Volumes, existingPod.Namespace, attachedVolumes); err != nil {
+			return false, nil, err
+		}
+	}
+
+	newVolumeCount := map[string]int{}
+	attachedVolumeCount := map[string]int{}
+
+	for volumeName, volumeLimitKey := range attachedVolumes {
+		if _, ok := newVolumes[volumeName]; ok {
+			delete(newVolumes, volumeName)
+		}
+		attachedVolumeCount[volumeLimitKey]++
+	}
+
+	for _, volumeLimitKey := range newVolumes {
+		newVolumeCount[volumeLimitKey]++
+	}
+
+	for volumeLimitKey, count := range newVolumeCount {
+		maxVolumeLimit, ok := nodeVolumeLimits[v1.ResourceName(volumeLimitKey)]
+		if ok {
+			currentVolumeCount := attachedVolumeCount[volumeLimitKey]
+			if currentVolumeCount+count > int(maxVolumeLimit) {
+				return false, []PredicateFailureReason{ErrMaxVolumeCountExceeded}, nil
+			}
+		}
+	}
+
+	return true, nil, nil
+}
+
+func (c *CSIMaxVolumeLimitChecker) filterAttachableVolumes(
+	volumes []v1.Volume, namespace string, result map[string]string) error {
+
+	for _, vol := range volumes {
+		// CSI volumes can only be used as persistent volumes
+		if vol.PersistentVolumeClaim == nil {
+			continue
+		}
+		pvcName := vol.PersistentVolumeClaim.ClaimName
+
+		if pvcName == "" {
+			return fmt.Errorf("PersistentVolumeClaim had no name")
+		}
+
+		pvc, err := c.pvcInfo.GetPersistentVolumeClaimInfo(namespace, pvcName)
+
+		if err != nil {
+			klog.V(4).Infof("Unable to look up PVC info for %s/%s", namespace, pvcName)
+			continue
+		}
+
+		pvName := pvc.Spec.VolumeName
+		// TODO - the actual handling of unbound PVCs will be fixed by late binding design.
+		if pvName == "" {
+			klog.V(4).Infof("Persistent volume had no name for claim %s/%s", namespace, pvcName)
+			continue
+		}
+		pv, err := c.pvInfo.GetPersistentVolumeInfo(pvName)
+
+		if err != nil {
+			klog.V(4).Infof("Unable to look up PV info for PVC %s/%s and PV %s", namespace, pvcName, pvName)
+			continue
+		}
+
+		csiSource := pv.Spec.PersistentVolumeSource.CSI
+		if csiSource == nil {
+			klog.V(4).Infof("Not considering non-CSI volume %s/%s", namespace, pvcName)
+			continue
+		}
+		driverName := csiSource.Driver
+		volumeLimitKey := volumeutil.GetCSIAttachLimitKey(driverName)
+		result[csiSource.VolumeHandle] = volumeLimitKey
+
+	}
+	return nil
+}

vendor/k8s.io/kubernetes/pkg/scheduler/algorithm/predicates/error.go

@@ -0,0 +1,155 @@
+/*
+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 predicates
+
+import (
+	"fmt"
+
+	"k8s.io/api/core/v1"
+)
+
+var (
+	// The predicateName tries to be consistent as the predicate name used in DefaultAlgorithmProvider defined in
+	// defaults.go (which tend to be stable for backward compatibility)
+
+	// NOTE: If you add a new predicate failure error for a predicate that can never
+	// be made to pass by removing pods, or you change an existing predicate so that
+	// it can never be made to pass by removing pods, you need to add the predicate
+	// failure error in nodesWherePreemptionMightHelp() in scheduler/core/generic_scheduler.go
+
+	// ErrDiskConflict is used for NoDiskConflict predicate error.
+	ErrDiskConflict = newPredicateFailureError("NoDiskConflict", "node(s) had no available disk")
+	// ErrVolumeZoneConflict is used for NoVolumeZoneConflict predicate error.
+	ErrVolumeZoneConflict = newPredicateFailureError("NoVolumeZoneConflict", "node(s) had no available volume zone")
+	// ErrNodeSelectorNotMatch is used for MatchNodeSelector predicate error.
+	ErrNodeSelectorNotMatch = newPredicateFailureError("MatchNodeSelector", "node(s) didn't match node selector")
+	// ErrPodAffinityNotMatch is used for MatchInterPodAffinity predicate error.
+	ErrPodAffinityNotMatch = newPredicateFailureError("MatchInterPodAffinity", "node(s) didn't match pod affinity/anti-affinity")
+	// ErrPodAffinityRulesNotMatch is used for PodAffinityRulesNotMatch predicate error.
+	ErrPodAffinityRulesNotMatch = newPredicateFailureError("PodAffinityRulesNotMatch", "node(s) didn't match pod affinity rules")
+	// ErrPodAntiAffinityRulesNotMatch is used for PodAntiAffinityRulesNotMatch predicate error.
+	ErrPodAntiAffinityRulesNotMatch = newPredicateFailureError("PodAntiAffinityRulesNotMatch", "node(s) didn't match pod anti-affinity rules")
+	// ErrExistingPodsAntiAffinityRulesNotMatch is used for ExistingPodsAntiAffinityRulesNotMatch predicate error.
+	ErrExistingPodsAntiAffinityRulesNotMatch = newPredicateFailureError("ExistingPodsAntiAffinityRulesNotMatch", "node(s) didn't satisfy existing pods anti-affinity rules")
+	// ErrTaintsTolerationsNotMatch is used for PodToleratesNodeTaints predicate error.
+	ErrTaintsTolerationsNotMatch = newPredicateFailureError("PodToleratesNodeTaints", "node(s) had taints that the pod didn't tolerate")
+	// ErrPodNotMatchHostName is used for HostName predicate error.
+	ErrPodNotMatchHostName = newPredicateFailureError("HostName", "node(s) didn't match the requested hostname")
+	// ErrPodNotFitsHostPorts is used for PodFitsHostPorts predicate error.
+	ErrPodNotFitsHostPorts = newPredicateFailureError("PodFitsHostPorts", "node(s) didn't have free ports for the requested pod ports")
+	// ErrNodeLabelPresenceViolated is used for CheckNodeLabelPresence predicate error.
+	ErrNodeLabelPresenceViolated = newPredicateFailureError("CheckNodeLabelPresence", "node(s) didn't have the requested labels")
+	// ErrServiceAffinityViolated is used for CheckServiceAffinity predicate error.
+	ErrServiceAffinityViolated = newPredicateFailureError("CheckServiceAffinity", "node(s) didn't match service affinity")
+	// ErrMaxVolumeCountExceeded is used for MaxVolumeCount predicate error.
+	ErrMaxVolumeCountExceeded = newPredicateFailureError("MaxVolumeCount", "node(s) exceed max volume count")
+	// ErrNodeUnderMemoryPressure is used for NodeUnderMemoryPressure predicate error.
+	ErrNodeUnderMemoryPressure = newPredicateFailureError("NodeUnderMemoryPressure", "node(s) had memory pressure")
+	// ErrNodeUnderDiskPressure is used for NodeUnderDiskPressure predicate error.
+	ErrNodeUnderDiskPressure = newPredicateFailureError("NodeUnderDiskPressure", "node(s) had disk pressure")
+	// ErrNodeUnderPIDPressure is used for NodeUnderPIDPressure predicate error.
+	ErrNodeUnderPIDPressure = newPredicateFailureError("NodeUnderPIDPressure", "node(s) had pid pressure")
+	// ErrNodeNotReady is used for NodeNotReady predicate error.
+	ErrNodeNotReady = newPredicateFailureError("NodeNotReady", "node(s) were not ready")
+	// ErrNodeNetworkUnavailable is used for NodeNetworkUnavailable predicate error.
+	ErrNodeNetworkUnavailable = newPredicateFailureError("NodeNetworkUnavailable", "node(s) had unavailable network")
+	// ErrNodeUnschedulable is used for NodeUnschedulable predicate error.
+	ErrNodeUnschedulable = newPredicateFailureError("NodeUnschedulable", "node(s) were unschedulable")
+	// ErrNodeUnknownCondition is used for NodeUnknownCondition predicate error.
+	ErrNodeUnknownCondition = newPredicateFailureError("NodeUnknownCondition", "node(s) had unknown conditions")
+	// ErrVolumeNodeConflict is used for VolumeNodeAffinityConflict predicate error.
+	ErrVolumeNodeConflict = newPredicateFailureError("VolumeNodeAffinityConflict", "node(s) had volume node affinity conflict")
+	// ErrVolumeBindConflict is used for VolumeBindingNoMatch predicate error.
+	ErrVolumeBindConflict = newPredicateFailureError("VolumeBindingNoMatch", "node(s) didn't find available persistent volumes to bind")
+	// ErrFakePredicate is used for test only. The fake predicates returning false also returns error
+	// as ErrFakePredicate.
+	ErrFakePredicate = newPredicateFailureError("FakePredicateError", "Nodes failed the fake predicate")
+)
+
+// InsufficientResourceError is an error type that indicates what kind of resource limit is
+// hit and caused the unfitting failure.
+type InsufficientResourceError struct {
+	// resourceName is the name of the resource that is insufficient
+	ResourceName v1.ResourceName
+	requested    int64
+	used         int64
+	capacity     int64
+}
+
+// NewInsufficientResourceError returns an InsufficientResourceError.
+func NewInsufficientResourceError(resourceName v1.ResourceName, requested, used, capacity int64) *InsufficientResourceError {
+	return &InsufficientResourceError{
+		ResourceName: resourceName,
+		requested:    requested,
+		used:         used,
+		capacity:     capacity,
+	}
+}
+
+func (e *InsufficientResourceError) Error() string {
+	return fmt.Sprintf("Node didn't have enough resource: %s, requested: %d, used: %d, capacity: %d",
+		e.ResourceName, e.requested, e.used, e.capacity)
+}
+
+// GetReason returns the reason of the InsufficientResourceError.
+func (e *InsufficientResourceError) GetReason() string {
+	return fmt.Sprintf("Insufficient %v", e.ResourceName)
+}
+
+// GetInsufficientAmount returns the amount of the insufficient resource of the error.
+func (e *InsufficientResourceError) GetInsufficientAmount() int64 {
+	return e.requested - (e.capacity - e.used)
+}
+
+// PredicateFailureError describes a failure error of predicate.
+type PredicateFailureError struct {
+	PredicateName string
+	PredicateDesc string
+}
+
+func newPredicateFailureError(predicateName, predicateDesc string) *PredicateFailureError {
+	return &PredicateFailureError{PredicateName: predicateName, PredicateDesc: predicateDesc}
+}
+
+func (e *PredicateFailureError) Error() string {
+	return fmt.Sprintf("Predicate %s failed", e.PredicateName)
+}
+
+// GetReason returns the reason of the PredicateFailureError.
+func (e *PredicateFailureError) GetReason() string {
+	return e.PredicateDesc
+}
+
+// PredicateFailureReason interface represents the failure reason of a predicate.
+type PredicateFailureReason interface {
+	GetReason() string
+}
+
+// FailureReason describes a failure reason.
+type FailureReason struct {
+	reason string
+}
+
+// NewFailureReason creates a FailureReason with message.
+func NewFailureReason(msg string) *FailureReason {
+	return &FailureReason{reason: msg}
+}
+
+// GetReason returns the reason of the FailureReason.
+func (e *FailureReason) GetReason() string {
+	return e.reason
+}

vendor/k8s.io/kubernetes/pkg/scheduler/algorithm/predicates/metadata.go

@@ -0,0 +1,542 @@
+/*
+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 predicates
+
+import (
+	"context"
+	"fmt"
+	"sync"
+
+	"k8s.io/klog"
+
+	"k8s.io/api/core/v1"
+	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
+	"k8s.io/apimachinery/pkg/labels"
+	"k8s.io/apimachinery/pkg/util/sets"
+	"k8s.io/client-go/util/workqueue"
+	"k8s.io/kubernetes/pkg/scheduler/algorithm"
+	priorityutil "k8s.io/kubernetes/pkg/scheduler/algorithm/priorities/util"
+	schedulernodeinfo "k8s.io/kubernetes/pkg/scheduler/nodeinfo"
+	schedutil "k8s.io/kubernetes/pkg/scheduler/util"
+)
+
+// PredicateMetadata interface represents anything that can access a predicate metadata.
+type PredicateMetadata interface {
+	ShallowCopy() PredicateMetadata
+	AddPod(addedPod *v1.Pod, nodeInfo *schedulernodeinfo.NodeInfo) error
+	RemovePod(deletedPod *v1.Pod) error
+}
+
+// PredicateMetadataProducer is a function that computes predicate metadata for a given pod.
+type PredicateMetadataProducer func(pod *v1.Pod, nodeNameToInfo map[string]*schedulernodeinfo.NodeInfo) PredicateMetadata
+
+// PredicateMetadataFactory defines a factory of predicate metadata.
+type PredicateMetadataFactory struct {
+	podLister algorithm.PodLister
+}
+
+// AntiAffinityTerm's topology key value used in predicate metadata
+type topologyPair struct {
+	key   string
+	value string
+}
+
+//  Note that predicateMetadata and matchingPodAntiAffinityTerm need to be declared in the same file
+//  due to the way declarations are processed in predicate declaration unit tests.
+type matchingPodAntiAffinityTerm struct {
+	term *v1.PodAffinityTerm
+	node *v1.Node
+}
+
+type podSet map[*v1.Pod]struct{}
+
+type topologyPairSet map[topologyPair]struct{}
+
+// topologyPairsMaps keeps topologyPairToAntiAffinityPods and antiAffinityPodToTopologyPairs in sync
+// as they are the inverse of each others.
+type topologyPairsMaps struct {
+	topologyPairToPods map[topologyPair]podSet
+	podToTopologyPairs map[string]topologyPairSet
+}
+
+// NOTE: When new fields are added/removed or logic is changed, please make sure that
+// RemovePod, AddPod, and ShallowCopy functions are updated to work with the new changes.
+type predicateMetadata struct {
+	pod           *v1.Pod
+	podBestEffort bool
+	podRequest    *schedulernodeinfo.Resource
+	podPorts      []*v1.ContainerPort
+
+	topologyPairsAntiAffinityPodsMap *topologyPairsMaps
+	// A map of topology pairs to a list of Pods that can potentially match
+	// the affinity terms of the "pod" and its inverse.
+	topologyPairsPotentialAffinityPods *topologyPairsMaps
+	// A map of topology pairs to a list of Pods that can potentially match
+	// the anti-affinity terms of the "pod" and its inverse.
+	topologyPairsPotentialAntiAffinityPods *topologyPairsMaps
+	serviceAffinityInUse                   bool
+	serviceAffinityMatchingPodList         []*v1.Pod
+	serviceAffinityMatchingPodServices     []*v1.Service
+	// ignoredExtendedResources is a set of extended resource names that will
+	// be ignored in the PodFitsResources predicate.
+	//
+	// They can be scheduler extender managed resources, the consumption of
+	// which should be accounted only by the extenders. This set is synthesized
+	// from scheduler extender configuration and does not change per pod.
+	ignoredExtendedResources sets.String
+}
+
+// Ensure that predicateMetadata implements algorithm.PredicateMetadata.
+var _ PredicateMetadata = &predicateMetadata{}
+
+// predicateMetadataProducer function produces predicate metadata. It is stored in a global variable below
+// and used to modify the return values of PredicateMetadataProducer
+type predicateMetadataProducer func(pm *predicateMetadata)
+
+var predicateMetaProducerRegisterLock sync.Mutex
+var predicateMetadataProducers = make(map[string]predicateMetadataProducer)
+
+// RegisterPredicateMetadataProducer registers a PredicateMetadataProducer.
+func RegisterPredicateMetadataProducer(predicateName string, precomp predicateMetadataProducer) {
+	predicateMetaProducerRegisterLock.Lock()
+	defer predicateMetaProducerRegisterLock.Unlock()
+	predicateMetadataProducers[predicateName] = precomp
+}
+
+// EmptyPredicateMetadataProducer returns a no-op MetadataProducer type.
+func EmptyPredicateMetadataProducer(pod *v1.Pod, nodeNameToInfo map[string]*schedulernodeinfo.NodeInfo) PredicateMetadata {
+	return nil
+}
+
+// RegisterPredicateMetadataProducerWithExtendedResourceOptions registers a
+// PredicateMetadataProducer that creates predicate metadata with the provided
+// options for extended resources.
+//
+// See the comments in "predicateMetadata" for the explanation of the options.
+func RegisterPredicateMetadataProducerWithExtendedResourceOptions(ignoredExtendedResources sets.String) {
+	RegisterPredicateMetadataProducer("PredicateWithExtendedResourceOptions", func(pm *predicateMetadata) {
+		pm.ignoredExtendedResources = ignoredExtendedResources
+	})
+}
+
+// NewPredicateMetadataFactory creates a PredicateMetadataFactory.
+func NewPredicateMetadataFactory(podLister algorithm.PodLister) PredicateMetadataProducer {
+	factory := &PredicateMetadataFactory{
+		podLister,
+	}
+	return factory.GetMetadata
+}
+
+// GetMetadata returns the predicateMetadata used which will be used by various predicates.
+func (pfactory *PredicateMetadataFactory) GetMetadata(pod *v1.Pod, nodeNameToInfoMap map[string]*schedulernodeinfo.NodeInfo) PredicateMetadata {
+	// If we cannot compute metadata, just return nil
+	if pod == nil {
+		return nil
+	}
+	// existingPodAntiAffinityMap will be used later for efficient check on existing pods' anti-affinity
+	existingPodAntiAffinityMap, err := getTPMapMatchingExistingAntiAffinity(pod, nodeNameToInfoMap)
+	if err != nil {
+		return nil
+	}
+	// incomingPodAffinityMap will be used later for efficient check on incoming pod's affinity
+	// incomingPodAntiAffinityMap will be used later for efficient check on incoming pod's anti-affinity
+	incomingPodAffinityMap, incomingPodAntiAffinityMap, err := getTPMapMatchingIncomingAffinityAntiAffinity(pod, nodeNameToInfoMap)
+	if err != nil {
+		klog.Errorf("[predicate meta data generation] error finding pods that match affinity terms: %v", err)
+		return nil
+	}
+	predicateMetadata := &predicateMetadata{
+		pod:                                    pod,
+		podBestEffort:                          isPodBestEffort(pod),
+		podRequest:                             GetResourceRequest(pod),
+		podPorts:                               schedutil.GetContainerPorts(pod),
+		topologyPairsPotentialAffinityPods:     incomingPodAffinityMap,
+		topologyPairsPotentialAntiAffinityPods: incomingPodAntiAffinityMap,
+		topologyPairsAntiAffinityPodsMap:       existingPodAntiAffinityMap,
+	}
+	for predicateName, precomputeFunc := range predicateMetadataProducers {
+		klog.V(10).Infof("Precompute: %v", predicateName)
+		precomputeFunc(predicateMetadata)
+	}
+	return predicateMetadata
+}
+
+// returns a pointer to a new topologyPairsMaps
+func newTopologyPairsMaps() *topologyPairsMaps {
+	return &topologyPairsMaps{topologyPairToPods: make(map[topologyPair]podSet),
+		podToTopologyPairs: make(map[string]topologyPairSet)}
+}
+
+func (topologyPairsMaps *topologyPairsMaps) addTopologyPair(pair topologyPair, pod *v1.Pod) {
+	podFullName := schedutil.GetPodFullName(pod)
+	if topologyPairsMaps.topologyPairToPods[pair] == nil {
+		topologyPairsMaps.topologyPairToPods[pair] = make(map[*v1.Pod]struct{})
+	}
+	topologyPairsMaps.topologyPairToPods[pair][pod] = struct{}{}
+	if topologyPairsMaps.podToTopologyPairs[podFullName] == nil {
+		topologyPairsMaps.podToTopologyPairs[podFullName] = make(map[topologyPair]struct{})
+	}
+	topologyPairsMaps.podToTopologyPairs[podFullName][pair] = struct{}{}
+}
+
+func (topologyPairsMaps *topologyPairsMaps) removePod(deletedPod *v1.Pod) {
+	deletedPodFullName := schedutil.GetPodFullName(deletedPod)
+	for pair := range topologyPairsMaps.podToTopologyPairs[deletedPodFullName] {
+		delete(topologyPairsMaps.topologyPairToPods[pair], deletedPod)
+		if len(topologyPairsMaps.topologyPairToPods[pair]) == 0 {
+			delete(topologyPairsMaps.topologyPairToPods, pair)
+		}
+	}
+	delete(topologyPairsMaps.podToTopologyPairs, deletedPodFullName)
+}
+
+func (topologyPairsMaps *topologyPairsMaps) appendMaps(toAppend *topologyPairsMaps) {
+	if toAppend == nil {
+		return
+	}
+	for pair := range toAppend.topologyPairToPods {
+		for pod := range toAppend.topologyPairToPods[pair] {
+			topologyPairsMaps.addTopologyPair(pair, pod)
+		}
+	}
+}
+
+// RemovePod changes predicateMetadata assuming that the given `deletedPod` is
+// deleted from the system.
+func (meta *predicateMetadata) RemovePod(deletedPod *v1.Pod) error {
+	deletedPodFullName := schedutil.GetPodFullName(deletedPod)
+	if deletedPodFullName == schedutil.GetPodFullName(meta.pod) {
+		return fmt.Errorf("deletedPod and meta.pod must not be the same")
+	}
+	meta.topologyPairsAntiAffinityPodsMap.removePod(deletedPod)
+	// Delete pod from the matching affinity or anti-affinity topology pairs maps.
+	meta.topologyPairsPotentialAffinityPods.removePod(deletedPod)
+	meta.topologyPairsPotentialAntiAffinityPods.removePod(deletedPod)
+	// All pods in the serviceAffinityMatchingPodList are in the same namespace.
+	// So, if the namespace of the first one is not the same as the namespace of the
+	// deletedPod, we don't need to check the list, as deletedPod isn't in the list.
+	if meta.serviceAffinityInUse &&
+		len(meta.serviceAffinityMatchingPodList) > 0 &&
+		deletedPod.Namespace == meta.serviceAffinityMatchingPodList[0].Namespace {
+		for i, pod := range meta.serviceAffinityMatchingPodList {
+			if schedutil.GetPodFullName(pod) == deletedPodFullName {
+				meta.serviceAffinityMatchingPodList = append(
+					meta.serviceAffinityMatchingPodList[:i],
+					meta.serviceAffinityMatchingPodList[i+1:]...)
+				break
+			}
+		}
+	}
+	return nil
+}
+
+// AddPod changes predicateMetadata assuming that `newPod` is added to the
+// system.
+func (meta *predicateMetadata) AddPod(addedPod *v1.Pod, nodeInfo *schedulernodeinfo.NodeInfo) error {
+	addedPodFullName := schedutil.GetPodFullName(addedPod)
+	if addedPodFullName == schedutil.GetPodFullName(meta.pod) {
+		return fmt.Errorf("addedPod and meta.pod must not be the same")
+	}
+	if nodeInfo.Node() == nil {
+		return fmt.Errorf("invalid node in nodeInfo")
+	}
+	// Add matching anti-affinity terms of the addedPod to the map.
+	topologyPairsMaps, err := getMatchingAntiAffinityTopologyPairsOfPod(meta.pod, addedPod, nodeInfo.Node())
+	if err != nil {
+		return err
+	}
+	meta.topologyPairsAntiAffinityPodsMap.appendMaps(topologyPairsMaps)
+	// Add the pod to nodeNameToMatchingAffinityPods and nodeNameToMatchingAntiAffinityPods if needed.
+	affinity := meta.pod.Spec.Affinity
+	podNodeName := addedPod.Spec.NodeName
+	if affinity != nil && len(podNodeName) > 0 {
+		podNode := nodeInfo.Node()
+		// It is assumed that when the added pod matches affinity of the meta.pod, all the terms must match,
+		// this should be changed when the implementation of targetPodMatchesAffinityOfPod/podMatchesAffinityTermProperties
+		// is changed
+		if targetPodMatchesAffinityOfPod(meta.pod, addedPod) {
+			affinityTerms := GetPodAffinityTerms(affinity.PodAffinity)
+			for _, term := range affinityTerms {
+				if topologyValue, ok := podNode.Labels[term.TopologyKey]; ok {
+					pair := topologyPair{key: term.TopologyKey, value: topologyValue}
+					meta.topologyPairsPotentialAffinityPods.addTopologyPair(pair, addedPod)
+				}
+			}
+		}
+		if targetPodMatchesAntiAffinityOfPod(meta.pod, addedPod) {
+			antiAffinityTerms := GetPodAntiAffinityTerms(affinity.PodAntiAffinity)
+			for _, term := range antiAffinityTerms {
+				if topologyValue, ok := podNode.Labels[term.TopologyKey]; ok {
+					pair := topologyPair{key: term.TopologyKey, value: topologyValue}
+					meta.topologyPairsPotentialAntiAffinityPods.addTopologyPair(pair, addedPod)
+				}
+			}
+		}
+	}
+	// If addedPod is in the same namespace as the meta.pod, update the list
+	// of matching pods if applicable.
+	if meta.serviceAffinityInUse && addedPod.Namespace == meta.pod.Namespace {
+		selector := CreateSelectorFromLabels(meta.pod.Labels)
+		if selector.Matches(labels.Set(addedPod.Labels)) {
+			meta.serviceAffinityMatchingPodList = append(meta.serviceAffinityMatchingPodList,
+				addedPod)
+		}
+	}
+	return nil
+}
+
+// ShallowCopy copies a metadata struct into a new struct and creates a copy of
+// its maps and slices, but it does not copy the contents of pointer values.
+func (meta *predicateMetadata) ShallowCopy() PredicateMetadata {
+	newPredMeta := &predicateMetadata{
+		pod:                      meta.pod,
+		podBestEffort:            meta.podBestEffort,
+		podRequest:               meta.podRequest,
+		serviceAffinityInUse:     meta.serviceAffinityInUse,
+		ignoredExtendedResources: meta.ignoredExtendedResources,
+	}
+	newPredMeta.podPorts = append([]*v1.ContainerPort(nil), meta.podPorts...)
+	newPredMeta.topologyPairsPotentialAffinityPods = newTopologyPairsMaps()
+	newPredMeta.topologyPairsPotentialAffinityPods.appendMaps(meta.topologyPairsPotentialAffinityPods)
+	newPredMeta.topologyPairsPotentialAntiAffinityPods = newTopologyPairsMaps()
+	newPredMeta.topologyPairsPotentialAntiAffinityPods.appendMaps(meta.topologyPairsPotentialAntiAffinityPods)
+	newPredMeta.topologyPairsAntiAffinityPodsMap = newTopologyPairsMaps()
+	newPredMeta.topologyPairsAntiAffinityPodsMap.appendMaps(meta.topologyPairsAntiAffinityPodsMap)
+	newPredMeta.serviceAffinityMatchingPodServices = append([]*v1.Service(nil),
+		meta.serviceAffinityMatchingPodServices...)
+	newPredMeta.serviceAffinityMatchingPodList = append([]*v1.Pod(nil),
+		meta.serviceAffinityMatchingPodList...)
+	return (PredicateMetadata)(newPredMeta)
+}
+
+type affinityTermProperties struct {
+	namespaces sets.String
+	selector   labels.Selector
+}
+
+// getAffinityTermProperties receives a Pod and affinity terms and returns the namespaces and
+// selectors of the terms.
+func getAffinityTermProperties(pod *v1.Pod, terms []v1.PodAffinityTerm) (properties []*affinityTermProperties, err error) {
+	if terms == nil {
+		return properties, nil
+	}
+
+	for _, term := range terms {
+		namespaces := priorityutil.GetNamespacesFromPodAffinityTerm(pod, &term)
+		selector, err := metav1.LabelSelectorAsSelector(term.LabelSelector)
+		if err != nil {
+			return nil, err
+		}
+		properties = append(properties, &affinityTermProperties{namespaces: namespaces, selector: selector})
+	}
+	return properties, nil
+}
+
+// podMatchesAllAffinityTermProperties returns true IFF the given pod matches all the given properties.
+func podMatchesAllAffinityTermProperties(pod *v1.Pod, properties []*affinityTermProperties) bool {
+	if len(properties) == 0 {
+		return false
+	}
+	for _, property := range properties {
+		if !priorityutil.PodMatchesTermsNamespaceAndSelector(pod, property.namespaces, property.selector) {
+			return false
+		}
+	}
+	return true
+}
+
+// podMatchesAnyAffinityTermProperties returns true if the given pod matches any given property.
+func podMatchesAnyAffinityTermProperties(pod *v1.Pod, properties []*affinityTermProperties) bool {
+	if len(properties) == 0 {
+		return false
+	}
+	for _, property := range properties {
+		if priorityutil.PodMatchesTermsNamespaceAndSelector(pod, property.namespaces, property.selector) {
+			return true
+		}
+	}
+	return false
+}
+
+// getTPMapMatchingExistingAntiAffinity calculates the following for each existing pod on each node:
+// (1) Whether it has PodAntiAffinity
+// (2) Whether any AffinityTerm matches the incoming pod
+func getTPMapMatchingExistingAntiAffinity(pod *v1.Pod, nodeInfoMap map[string]*schedulernodeinfo.NodeInfo) (*topologyPairsMaps, error) {
+	allNodeNames := make([]string, 0, len(nodeInfoMap))
+	for name := range nodeInfoMap {
+		allNodeNames = append(allNodeNames, name)
+	}
+
+	var lock sync.Mutex
+	var firstError error
+
+	topologyMaps := newTopologyPairsMaps()
+
+	appendTopologyPairsMaps := func(toAppend *topologyPairsMaps) {
+		lock.Lock()
+		defer lock.Unlock()
+		topologyMaps.appendMaps(toAppend)
+	}
+	catchError := func(err error) {
+		lock.Lock()
+		defer lock.Unlock()
+		if firstError == nil {
+			firstError = err
+		}
+	}
+
+	processNode := func(i int) {
+		nodeInfo := nodeInfoMap[allNodeNames[i]]
+		node := nodeInfo.Node()
+		if node == nil {
+			catchError(fmt.Errorf("node not found"))
+			return
+		}
+		for _, existingPod := range nodeInfo.PodsWithAffinity() {
+			existingPodTopologyMaps, err := getMatchingAntiAffinityTopologyPairsOfPod(pod, existingPod, node)
+			if err != nil {
+				catchError(err)
+				return
+			}
+			appendTopologyPairsMaps(existingPodTopologyMaps)
+		}
+	}
+	workqueue.ParallelizeUntil(context.TODO(), 16, len(allNodeNames), processNode)
+	return topologyMaps, firstError
+}
+
+// getTPMapMatchingIncomingAffinityAntiAffinity finds existing Pods that match affinity terms of the given "pod".
+// It returns a topologyPairsMaps that are checked later by the affinity
+// predicate. With this topologyPairsMaps available, the affinity predicate does not
+// need to check all the pods in the cluster.
+func getTPMapMatchingIncomingAffinityAntiAffinity(pod *v1.Pod, nodeInfoMap map[string]*schedulernodeinfo.NodeInfo) (topologyPairsAffinityPodsMaps *topologyPairsMaps, topologyPairsAntiAffinityPodsMaps *topologyPairsMaps, err error) {
+	affinity := pod.Spec.Affinity
+	if affinity == nil || (affinity.PodAffinity == nil && affinity.PodAntiAffinity == nil) {
+		return newTopologyPairsMaps(), newTopologyPairsMaps(), nil
+	}
+
+	allNodeNames := make([]string, 0, len(nodeInfoMap))
+	for name := range nodeInfoMap {
+		allNodeNames = append(allNodeNames, name)
+	}
+
+	var lock sync.Mutex
+	var firstError error
+	topologyPairsAffinityPodsMaps = newTopologyPairsMaps()
+	topologyPairsAntiAffinityPodsMaps = newTopologyPairsMaps()
+	appendResult := func(nodeName string, nodeTopologyPairsAffinityPodsMaps, nodeTopologyPairsAntiAffinityPodsMaps *topologyPairsMaps) {
+		lock.Lock()
+		defer lock.Unlock()
+		if len(nodeTopologyPairsAffinityPodsMaps.topologyPairToPods) > 0 {
+			topologyPairsAffinityPodsMaps.appendMaps(nodeTopologyPairsAffinityPodsMaps)
+		}
+		if len(nodeTopologyPairsAntiAffinityPodsMaps.topologyPairToPods) > 0 {
+			topologyPairsAntiAffinityPodsMaps.appendMaps(nodeTopologyPairsAntiAffinityPodsMaps)
+		}
+	}
+
+	catchError := func(err error) {
+		lock.Lock()
+		defer lock.Unlock()
+		if firstError == nil {
+			firstError = err
+		}
+	}
+
+	affinityTerms := GetPodAffinityTerms(affinity.PodAffinity)
+	affinityProperties, err := getAffinityTermProperties(pod, affinityTerms)
+	if err != nil {
+		return nil, nil, err
+	}
+	antiAffinityTerms := GetPodAntiAffinityTerms(affinity.PodAntiAffinity)
+
+	processNode := func(i int) {
+		nodeInfo := nodeInfoMap[allNodeNames[i]]
+		node := nodeInfo.Node()
+		if node == nil {
+			catchError(fmt.Errorf("nodeInfo.Node is nil"))
+			return
+		}
+		nodeTopologyPairsAffinityPodsMaps := newTopologyPairsMaps()
+		nodeTopologyPairsAntiAffinityPodsMaps := newTopologyPairsMaps()
+		for _, existingPod := range nodeInfo.Pods() {
+			// Check affinity properties.
+			if podMatchesAllAffinityTermProperties(existingPod, affinityProperties) {
+				for _, term := range affinityTerms {
+					if topologyValue, ok := node.Labels[term.TopologyKey]; ok {
+						pair := topologyPair{key: term.TopologyKey, value: topologyValue}
+						nodeTopologyPairsAffinityPodsMaps.addTopologyPair(pair, existingPod)
+					}
+				}
+			}
+			// Check anti-affinity properties.
+			for _, term := range antiAffinityTerms {
+				namespaces := priorityutil.GetNamespacesFromPodAffinityTerm(pod, &term)
+				selector, err := metav1.LabelSelectorAsSelector(term.LabelSelector)
+				if err != nil {
+					catchError(err)
+					return
+				}
+				if priorityutil.PodMatchesTermsNamespaceAndSelector(existingPod, namespaces, selector) {
+					if topologyValue, ok := node.Labels[term.TopologyKey]; ok {
+						pair := topologyPair{key: term.TopologyKey, value: topologyValue}
+						nodeTopologyPairsAntiAffinityPodsMaps.addTopologyPair(pair, existingPod)
+					}
+				}
+			}
+		}
+		if len(nodeTopologyPairsAffinityPodsMaps.topologyPairToPods) > 0 || len(nodeTopologyPairsAntiAffinityPodsMaps.topologyPairToPods) > 0 {
+			appendResult(node.Name, nodeTopologyPairsAffinityPodsMaps, nodeTopologyPairsAntiAffinityPodsMaps)
+		}
+	}
+	workqueue.ParallelizeUntil(context.TODO(), 16, len(allNodeNames), processNode)
+	return topologyPairsAffinityPodsMaps, topologyPairsAntiAffinityPodsMaps, firstError
+}
+
+// targetPodMatchesAffinityOfPod returns true if "targetPod" matches ALL affinity terms of
+// "pod". This function does not check topology.
+// So, whether the targetPod actually matches or not needs further checks for a specific
+// node.
+func targetPodMatchesAffinityOfPod(pod, targetPod *v1.Pod) bool {
+	affinity := pod.Spec.Affinity
+	if affinity == nil || affinity.PodAffinity == nil {
+		return false
+	}
+	affinityProperties, err := getAffinityTermProperties(pod, GetPodAffinityTerms(affinity.PodAffinity))
+	if err != nil {
+		klog.Errorf("error in getting affinity properties of Pod %v", pod.Name)
+		return false
+	}
+	return podMatchesAllAffinityTermProperties(targetPod, affinityProperties)
+}
+
+// targetPodMatchesAntiAffinityOfPod returns true if "targetPod" matches ANY anti-affinity
+// term of "pod". This function does not check topology.
+// So, whether the targetPod actually matches or not needs further checks for a specific
+// node.
+func targetPodMatchesAntiAffinityOfPod(pod, targetPod *v1.Pod) bool {
+	affinity := pod.Spec.Affinity
+	if affinity == nil || affinity.PodAntiAffinity == nil {
+		return false
+	}
+	properties, err := getAffinityTermProperties(pod, GetPodAntiAffinityTerms(affinity.PodAntiAffinity))
+	if err != nil {
+		klog.Errorf("error in getting anti-affinity properties of Pod %v", pod.Name)
+		return false
+	}
+	return podMatchesAnyAffinityTermProperties(targetPod, properties)
+}

vendor/k8s.io/kubernetes/pkg/scheduler/algorithm/predicates/testing_helper.go

@@ -0,0 +1,85 @@
+/*
+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 predicates
+
+import (
+	"fmt"
+
+	"k8s.io/api/core/v1"
+	storagev1 "k8s.io/api/storage/v1"
+)
+
+// FakePersistentVolumeClaimInfo declares a []v1.PersistentVolumeClaim type for testing.
+type FakePersistentVolumeClaimInfo []v1.PersistentVolumeClaim
+
+// GetPersistentVolumeClaimInfo gets PVC matching the namespace and PVC ID.
+func (pvcs FakePersistentVolumeClaimInfo) GetPersistentVolumeClaimInfo(namespace string, pvcID string) (*v1.PersistentVolumeClaim, error) {
+	for _, pvc := range pvcs {
+		if pvc.Name == pvcID && pvc.Namespace == namespace {
+			return &pvc, nil
+		}
+	}
+	return nil, fmt.Errorf("Unable to find persistent volume claim: %s/%s", namespace, pvcID)
+}
+
+// FakeNodeInfo declares a v1.Node type for testing.
+type FakeNodeInfo v1.Node
+
+// GetNodeInfo return a fake node info object.
+func (n FakeNodeInfo) GetNodeInfo(nodeName string) (*v1.Node, error) {
+	node := v1.Node(n)
+	return &node, nil
+}
+
+// FakeNodeListInfo declares a []v1.Node type for testing.
+type FakeNodeListInfo []v1.Node
+
+// GetNodeInfo returns a fake node object in the fake nodes.
+func (nodes FakeNodeListInfo) GetNodeInfo(nodeName string) (*v1.Node, error) {
+	for _, node := range nodes {
+		if node.Name == nodeName {
+			return &node, nil
+		}
+	}
+	return nil, fmt.Errorf("Unable to find node: %s", nodeName)
+}
+
+// FakePersistentVolumeInfo declares a []v1.PersistentVolume type for testing.
+type FakePersistentVolumeInfo []v1.PersistentVolume
+
+// GetPersistentVolumeInfo returns a fake PV object in the fake PVs by PV ID.
+func (pvs FakePersistentVolumeInfo) GetPersistentVolumeInfo(pvID string) (*v1.PersistentVolume, error) {
+	for _, pv := range pvs {
+		if pv.Name == pvID {
+			return &pv, nil
+		}
+	}
+	return nil, fmt.Errorf("Unable to find persistent volume: %s", pvID)
+}
+
+// FakeStorageClassInfo declares a []storagev1.StorageClass type for testing.
+type FakeStorageClassInfo []storagev1.StorageClass
+
+// GetStorageClassInfo returns a fake storage class object in the fake storage classes by name.
+func (classes FakeStorageClassInfo) GetStorageClassInfo(name string) (*storagev1.StorageClass, error) {
+	for _, sc := range classes {
+		if sc.Name == name {
+			return &sc, nil
+		}
+	}
+	return nil, fmt.Errorf("Unable to find storage class: %s", name)
+}

vendor/k8s.io/kubernetes/pkg/scheduler/algorithm/predicates/utils.go

@@ -0,0 +1,79 @@
+/*
+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 predicates
+
+import (
+	"k8s.io/api/core/v1"
+	"k8s.io/apimachinery/pkg/labels"
+	schedulernodeinfo "k8s.io/kubernetes/pkg/scheduler/nodeinfo"
+)
+
+// FindLabelsInSet gets as many key/value pairs as possible out of a label set.
+func FindLabelsInSet(labelsToKeep []string, selector labels.Set) map[string]string {
+	aL := make(map[string]string)
+	for _, l := range labelsToKeep {
+		if selector.Has(l) {
+			aL[l] = selector.Get(l)
+		}
+	}
+	return aL
+}
+
+// AddUnsetLabelsToMap backfills missing values with values we find in a map.
+func AddUnsetLabelsToMap(aL map[string]string, labelsToAdd []string, labelSet labels.Set) {
+	for _, l := range labelsToAdd {
+		// if the label is already there, dont overwrite it.
+		if _, exists := aL[l]; exists {
+			continue
+		}
+		// otherwise, backfill this label.
+		if labelSet.Has(l) {
+			aL[l] = labelSet.Get(l)
+		}
+	}
+}
+
+// FilterPodsByNamespace filters pods outside a namespace from the given list.
+func FilterPodsByNamespace(pods []*v1.Pod, ns string) []*v1.Pod {
+	filtered := []*v1.Pod{}
+	for _, nsPod := range pods {
+		if nsPod.Namespace == ns {
+			filtered = append(filtered, nsPod)
+		}
+	}
+	return filtered
+}
+
+// CreateSelectorFromLabels is used to define a selector that corresponds to the keys in a map.
+func CreateSelectorFromLabels(aL map[string]string) labels.Selector {
+	if aL == nil || len(aL) == 0 {
+		return labels.Everything()
+	}
+	return labels.Set(aL).AsSelector()
+}
+
+// portsConflict check whether existingPorts and wantPorts conflict with each other
+// return true if we have a conflict
+func portsConflict(existingPorts schedulernodeinfo.HostPortInfo, wantPorts []*v1.ContainerPort) bool {
+	for _, cp := range wantPorts {
+		if existingPorts.CheckConflict(cp.HostIP, string(cp.Protocol), cp.HostPort) {
+			return true
+		}
+	}
+
+	return false
+}

vendor/k8s.io/kubernetes/pkg/scheduler/algorithm/priorities/util/non_zero.go

@@ -0,0 +1,52 @@
+/*
+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 "k8s.io/api/core/v1"
+
+// For each of these resources, a pod 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 machine 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.
+
+// DefaultMilliCPURequest defines default milli cpu request number.
+const DefaultMilliCPURequest int64 = 100 // 0.1 core
+// DefaultMemoryRequest defines default memory request size.
+const DefaultMemoryRequest int64 = 200 * 1024 * 1024 // 200 MB
+
+// GetNonzeroRequests returns the default resource request if none is found or
+// what is provided on the request.
+func GetNonzeroRequests(requests *v1.ResourceList) (int64, int64) {
+	var outMilliCPU, outMemory int64
+	// Override if un-set, but not if explicitly set to zero
+	if _, found := (*requests)[v1.ResourceCPU]; !found {
+		outMilliCPU = DefaultMilliCPURequest
+	} else {
+		outMilliCPU = requests.Cpu().MilliValue()
+	}
+	// Override if un-set, but not if explicitly set to zero
+	if _, found := (*requests)[v1.ResourceMemory]; !found {
+		outMemory = DefaultMemoryRequest
+	} else {
+		outMemory = requests.Memory().Value()
+	}
+	return outMilliCPU, outMemory
+}

vendor/k8s.io/kubernetes/pkg/scheduler/algorithm/priorities/util/topologies.go

@@ -0,0 +1,81 @@
+/*
+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 (
+	"k8s.io/api/core/v1"
+	"k8s.io/apimachinery/pkg/labels"
+	"k8s.io/apimachinery/pkg/util/sets"
+)
+
+// GetNamespacesFromPodAffinityTerm returns a set of names
+// according to the namespaces indicated in podAffinityTerm.
+// If namespaces is empty it considers the given pod's namespace.
+func GetNamespacesFromPodAffinityTerm(pod *v1.Pod, podAffinityTerm *v1.PodAffinityTerm) sets.String {
+	names := sets.String{}
+	if len(podAffinityTerm.Namespaces) == 0 {
+		names.Insert(pod.Namespace)
+	} else {
+		names.Insert(podAffinityTerm.Namespaces...)
+	}
+	return names
+}
+
+// PodMatchesTermsNamespaceAndSelector returns true if the given <pod>
+// matches the namespace and selector defined by <affinityPod>`s <term>.
+func PodMatchesTermsNamespaceAndSelector(pod *v1.Pod, namespaces sets.String, selector labels.Selector) bool {
+	if !namespaces.Has(pod.Namespace) {
+		return false
+	}
+
+	if !selector.Matches(labels.Set(pod.Labels)) {
+		return false
+	}
+	return true
+}
+
+// NodesHaveSameTopologyKey checks if nodeA and nodeB have same label value with given topologyKey as label key.
+// Returns false if topologyKey is empty.
+func NodesHaveSameTopologyKey(nodeA, nodeB *v1.Node, topologyKey string) bool {
+	if len(topologyKey) == 0 {
+		return false
+	}
+
+	if nodeA.Labels == nil || nodeB.Labels == nil {
+		return false
+	}
+
+	nodeALabel, okA := nodeA.Labels[topologyKey]
+	nodeBLabel, okB := nodeB.Labels[topologyKey]
+
+	// If found label in both nodes, check the label
+	if okB && okA {
+		return nodeALabel == nodeBLabel
+	}
+
+	return false
+}
+
+// Topologies contains topologies information of nodes.
+type Topologies struct {
+	DefaultKeys []string
+}
+
+// NodesHaveSameTopologyKey checks if nodeA and nodeB have same label value with given topologyKey as label key.
+func (tps *Topologies) NodesHaveSameTopologyKey(nodeA, nodeB *v1.Node, topologyKey string) bool {
+	return NodesHaveSameTopologyKey(nodeA, nodeB, topologyKey)
+}

vendor/k8s.io/kubernetes/pkg/scheduler/algorithm/scheduler_interface.go

@@ -0,0 +1,74 @@
+/*
+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 algorithm
+
+import (
+	"k8s.io/api/core/v1"
+	schedulerapi "k8s.io/kubernetes/pkg/scheduler/api"
+	schedulernodeinfo "k8s.io/kubernetes/pkg/scheduler/nodeinfo"
+)
+
+// SchedulerExtender is an interface for external processes to influence scheduling
+// decisions made by Kubernetes. This is typically needed for resources not directly
+// managed by Kubernetes.
+type SchedulerExtender interface {
+	// Name returns a unique name that identifies the extender.
+	Name() string
+
+	// Filter based on extender-implemented predicate functions. The filtered list is
+	// expected to be a subset of the supplied list. failedNodesMap optionally contains
+	// the list of failed nodes and failure reasons.
+	Filter(pod *v1.Pod,
+		nodes []*v1.Node, nodeNameToInfo map[string]*schedulernodeinfo.NodeInfo,
+	) (filteredNodes []*v1.Node, failedNodesMap schedulerapi.FailedNodesMap, err error)
+
+	// Prioritize based on extender-implemented priority functions. The returned scores & weight
+	// are used to compute the weighted score for an extender. The weighted scores are added to
+	// the scores computed  by Kubernetes scheduler. The total scores are used to do the host selection.
+	Prioritize(pod *v1.Pod, nodes []*v1.Node) (hostPriorities *schedulerapi.HostPriorityList, weight int, err error)
+
+	// Bind delegates the action of binding a pod to a node to the extender.
+	Bind(binding *v1.Binding) error
+
+	// IsBinder returns whether this extender is configured for the Bind method.
+	IsBinder() bool
+
+	// IsInterested returns true if at least one extended resource requested by
+	// this pod is managed by this extender.
+	IsInterested(pod *v1.Pod) bool
+
+	// ProcessPreemption returns nodes with their victim pods processed by extender based on
+	// given:
+	//   1. Pod to schedule
+	//   2. Candidate nodes and victim pods (nodeToVictims) generated by previous scheduling process.
+	//   3. nodeNameToInfo to restore v1.Node from node name if extender cache is enabled.
+	// The possible changes made by extender may include:
+	//   1. Subset of given candidate nodes after preemption phase of extender.
+	//   2. A different set of victim pod for every given candidate node after preemption phase of extender.
+	ProcessPreemption(
+		pod *v1.Pod,
+		nodeToVictims map[*v1.Node]*schedulerapi.Victims,
+		nodeNameToInfo map[string]*schedulernodeinfo.NodeInfo,
+	) (map[*v1.Node]*schedulerapi.Victims, error)
+
+	// SupportsPreemption returns if the scheduler extender support preemption or not.
+	SupportsPreemption() bool
+
+	// IsIgnorable returns true indicates scheduling should not fail when this extender
+	// is unavailable. This gives scheduler ability to fail fast and tolerate non-critical extenders as well.
+	IsIgnorable() bool
+}

vendor/k8s.io/kubernetes/pkg/scheduler/algorithm/types.go

@@ -0,0 +1,120 @@
+/*
+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 algorithm
+
+import (
+	apps "k8s.io/api/apps/v1"
+	"k8s.io/api/core/v1"
+	policyv1beta1 "k8s.io/api/policy/v1beta1"
+	"k8s.io/apimachinery/pkg/labels"
+	schedulerapi "k8s.io/kubernetes/pkg/scheduler/api"
+)
+
+// NodeFieldSelectorKeys is a map that: the key are node field selector keys; the values are
+// the functions to get the value of the node field.
+var NodeFieldSelectorKeys = map[string]func(*v1.Node) string{
+	schedulerapi.NodeFieldSelectorKeyNodeName: func(n *v1.Node) string { return n.Name },
+}
+
+// NodeLister interface represents anything that can list nodes for a scheduler.
+type NodeLister interface {
+	// We explicitly return []*v1.Node, instead of v1.NodeList, to avoid
+	// performing expensive copies that are unneeded.
+	List() ([]*v1.Node, error)
+}
+
+// PodFilter is a function to filter a pod. If pod passed return true else return false.
+type PodFilter func(*v1.Pod) bool
+
+// PodLister interface represents anything that can list pods for a scheduler.
+type PodLister interface {
+	// We explicitly return []*v1.Pod, instead of v1.PodList, to avoid
+	// performing expensive copies that are unneeded.
+	List(labels.Selector) ([]*v1.Pod, error)
+	// This is similar to "List()", but the returned slice does not
+	// contain pods that don't pass `podFilter`.
+	FilteredList(podFilter PodFilter, selector labels.Selector) ([]*v1.Pod, error)
+}
+
+// ServiceLister interface represents anything that can produce a list of services; the list is consumed by a scheduler.
+type ServiceLister interface {
+	// Lists all the services
+	List(labels.Selector) ([]*v1.Service, error)
+	// Gets the services for the given pod
+	GetPodServices(*v1.Pod) ([]*v1.Service, error)
+}
+
+// ControllerLister interface represents anything that can produce a list of ReplicationController; the list is consumed by a scheduler.
+type ControllerLister interface {
+	// Lists all the replication controllers
+	List(labels.Selector) ([]*v1.ReplicationController, error)
+	// Gets the services for the given pod
+	GetPodControllers(*v1.Pod) ([]*v1.ReplicationController, error)
+}
+
+// ReplicaSetLister interface represents anything that can produce a list of ReplicaSet; the list is consumed by a scheduler.
+type ReplicaSetLister interface {
+	// Gets the replicasets for the given pod
+	GetPodReplicaSets(*v1.Pod) ([]*apps.ReplicaSet, error)
+}
+
+// PDBLister interface represents anything that can list PodDisruptionBudget objects.
+type PDBLister interface {
+	// List() returns a list of PodDisruptionBudgets matching the selector.
+	List(labels.Selector) ([]*policyv1beta1.PodDisruptionBudget, error)
+}
+
+var _ ControllerLister = &EmptyControllerLister{}
+
+// EmptyControllerLister implements ControllerLister on []v1.ReplicationController returning empty data
+type EmptyControllerLister struct{}
+
+// List returns nil
+func (f EmptyControllerLister) List(labels.Selector) ([]*v1.ReplicationController, error) {
+	return nil, nil
+}
+
+// GetPodControllers returns nil
+func (f EmptyControllerLister) GetPodControllers(pod *v1.Pod) (controllers []*v1.ReplicationController, err error) {
+	return nil, nil
+}
+
+var _ ReplicaSetLister = &EmptyReplicaSetLister{}
+
+// EmptyReplicaSetLister implements ReplicaSetLister on []extensions.ReplicaSet returning empty data
+type EmptyReplicaSetLister struct{}
+
+// GetPodReplicaSets returns nil
+func (f EmptyReplicaSetLister) GetPodReplicaSets(pod *v1.Pod) (rss []*apps.ReplicaSet, err error) {
+	return nil, nil
+}
+
+// StatefulSetLister interface represents anything that can produce a list of StatefulSet; the list is consumed by a scheduler.
+type StatefulSetLister interface {
+	// Gets the StatefulSet for the given pod.
+	GetPodStatefulSets(*v1.Pod) ([]*apps.StatefulSet, error)
+}
+
+var _ StatefulSetLister = &EmptyStatefulSetLister{}
+
+// EmptyStatefulSetLister implements StatefulSetLister on []apps.StatefulSet returning empty data.
+type EmptyStatefulSetLister struct{}
+
+// GetPodStatefulSets of EmptyStatefulSetLister returns nil.
+func (f EmptyStatefulSetLister) GetPodStatefulSets(pod *v1.Pod) (sss []*apps.StatefulSet, err error) {
+	return nil, nil
+}

vendor/k8s.io/kubernetes/pkg/scheduler/api/doc.go

@@ -0,0 +1,20 @@
+/*
+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.
+*/
+
+// +k8s:deepcopy-gen=package
+
+// Package api contains scheduler API objects.
+package api // import "k8s.io/kubernetes/pkg/scheduler/api"

vendor/k8s.io/kubernetes/pkg/scheduler/api/register.go

@@ -0,0 +1,55 @@
+/*
+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 api
+
+import (
+	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
+	"k8s.io/apimachinery/pkg/runtime"
+	"k8s.io/apimachinery/pkg/runtime/schema"
+)
+
+// Scheme is the default instance of runtime.Scheme to which types in the Kubernetes API are already registered.
+// TODO: remove this, scheduler should not have its own scheme.
+var Scheme = runtime.NewScheme()
+
+// SchemeGroupVersion is group version used to register these objects
+// TODO this should be in the "scheduler" group
+var SchemeGroupVersion = schema.GroupVersion{Group: "", Version: runtime.APIVersionInternal}
+
+var (
+	// SchemeBuilder defines a SchemeBuilder object.
+	SchemeBuilder = runtime.NewSchemeBuilder(addKnownTypes)
+	// AddToScheme is used to add stored functions to scheme.
+	AddToScheme = SchemeBuilder.AddToScheme
+)
+
+func init() {
+	if err := addKnownTypes(Scheme); err != nil {
+		// Programmer error.
+		panic(err)
+	}
+}
+
+func addKnownTypes(scheme *runtime.Scheme) error {
+	if err := scheme.AddIgnoredConversionType(&metav1.TypeMeta{}, &metav1.TypeMeta{}); err != nil {
+		return err
+	}
+	scheme.AddKnownTypes(SchemeGroupVersion,
+		&Policy{},
+	)
+	return nil
+}

vendor/k8s.io/kubernetes/pkg/scheduler/api/types.go

@@ -0,0 +1,354 @@
+/*
+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 api
+
+import (
+	"time"
+
+	"k8s.io/api/core/v1"
+	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
+	"k8s.io/apimachinery/pkg/types"
+)
+
+const (
+	// MaxUint defines the max unsigned int value.
+	MaxUint = ^uint(0)
+	// MaxInt defines the max signed int value.
+	MaxInt = int(MaxUint >> 1)
+	// MaxTotalPriority defines the max total priority value.
+	MaxTotalPriority = MaxInt
+	// MaxPriority defines the max priority value.
+	MaxPriority = 10
+	// MaxWeight defines the max weight value.
+	MaxWeight = MaxInt / MaxPriority
+	// DefaultPercentageOfNodesToScore defines the percentage of nodes of all nodes
+	// that once found feasible, the scheduler stops looking for more nodes.
+	DefaultPercentageOfNodesToScore = 50
+)
+
+// +k8s:deepcopy-gen:interfaces=k8s.io/apimachinery/pkg/runtime.Object
+
+// Policy describes a struct of a policy resource in api.
+type Policy struct {
+	metav1.TypeMeta
+	// Holds the information to configure the fit predicate functions.
+	// If unspecified, the default predicate functions will be applied.
+	// If empty list, all predicates (except the mandatory ones) will be
+	// bypassed.
+	Predicates []PredicatePolicy
+	// Holds the information to configure the priority functions.
+	// If unspecified, the default priority functions will be applied.
+	// If empty list, all priority functions will be bypassed.
+	Priorities []PriorityPolicy
+	// Holds the information to communicate with the extender(s)
+	ExtenderConfigs []ExtenderConfig
+	// RequiredDuringScheduling affinity is not symmetric, but there is an implicit PreferredDuringScheduling affinity rule
+	// corresponding to every RequiredDuringScheduling affinity rule.
+	// HardPodAffinitySymmetricWeight represents the weight of implicit PreferredDuringScheduling affinity rule, in the range 1-100.
+	HardPodAffinitySymmetricWeight int32
+
+	// When AlwaysCheckAllPredicates is set to true, scheduler checks all
+	// the configured predicates even after one or more of them fails.
+	// When the flag is set to false, scheduler skips checking the rest
+	// of the predicates after it finds one predicate that failed.
+	AlwaysCheckAllPredicates bool
+}
+
+// PredicatePolicy describes a struct of a predicate policy.
+type PredicatePolicy struct {
+	// Identifier of the predicate policy
+	// For a custom predicate, the name can be user-defined
+	// For the Kubernetes provided predicates, the name is the identifier of the pre-defined predicate
+	Name string
+	// Holds the parameters to configure the given predicate
+	Argument *PredicateArgument
+}
+
+// PriorityPolicy describes a struct of a priority policy.
+type PriorityPolicy struct {
+	// Identifier of the priority policy
+	// For a custom priority, the name can be user-defined
+	// For the Kubernetes provided priority functions, the name is the identifier of the pre-defined priority function
+	Name string
+	// The numeric multiplier for the node scores that the priority function generates
+	// The weight should be a positive integer
+	Weight int
+	// Holds the parameters to configure the given priority function
+	Argument *PriorityArgument
+}
+
+// PredicateArgument represents the arguments to configure predicate functions in scheduler policy configuration.
+// Only one of its members may be specified
+type PredicateArgument struct {
+	// The predicate that provides affinity for pods belonging to a service
+	// It uses a label to identify nodes that belong to the same "group"
+	ServiceAffinity *ServiceAffinity
+	// The predicate that checks whether a particular node has a certain label
+	// defined or not, regardless of value
+	LabelsPresence *LabelsPresence
+}
+
+// PriorityArgument represents the arguments to configure priority functions in scheduler policy configuration.
+// Only one of its members may be specified
+type PriorityArgument struct {
+	// The priority function that ensures a good spread (anti-affinity) for pods belonging to a service
+	// It uses a label to identify nodes that belong to the same "group"
+	ServiceAntiAffinity *ServiceAntiAffinity
+	// The priority function that checks whether a particular node has a certain label
+	// defined or not, regardless of value
+	LabelPreference *LabelPreference
+	// The RequestedToCapacityRatio priority function is parametrized with function shape.
+	RequestedToCapacityRatioArguments *RequestedToCapacityRatioArguments
+}
+
+// ServiceAffinity holds the parameters that are used to configure the corresponding predicate in scheduler policy configuration.
+type ServiceAffinity struct {
+	// The list of labels that identify node "groups"
+	// All of the labels should match for the node to be considered a fit for hosting the pod
+	Labels []string
+}
+
+// LabelsPresence holds the parameters that are used to configure the corresponding predicate in scheduler policy configuration.
+type LabelsPresence struct {
+	// The list of labels that identify node "groups"
+	// All of the labels should be either present (or absent) for the node to be considered a fit for hosting the pod
+	Labels []string
+	// The boolean flag that indicates whether the labels should be present or absent from the node
+	Presence bool
+}
+
+// ServiceAntiAffinity holds the parameters that are used to configure the corresponding priority function
+type ServiceAntiAffinity struct {
+	// Used to identify node "groups"
+	Label string
+}
+
+// LabelPreference holds the parameters that are used to configure the corresponding priority function
+type LabelPreference struct {
+	// Used to identify node "groups"
+	Label string
+	// This is a boolean flag
+	// If true, higher priority is given to nodes that have the label
+	// If false, higher priority is given to nodes that do not have the label
+	Presence bool
+}
+
+// RequestedToCapacityRatioArguments holds arguments specific to RequestedToCapacityRatio priority function
+type RequestedToCapacityRatioArguments struct {
+	// Array of point defining priority function shape
+	UtilizationShape []UtilizationShapePoint
+}
+
+// UtilizationShapePoint represents single point of priority function shape
+type UtilizationShapePoint struct {
+	// Utilization (x axis). Valid values are 0 to 100. Fully utilized node maps to 100.
+	Utilization int
+	// Score assigned to given utilization (y axis). Valid values are 0 to 10.
+	Score int
+}
+
+// ExtenderManagedResource describes the arguments of extended resources
+// managed by an extender.
+type ExtenderManagedResource struct {
+	// Name is the extended resource name.
+	Name v1.ResourceName
+	// IgnoredByScheduler indicates whether kube-scheduler should ignore this
+	// resource when applying predicates.
+	IgnoredByScheduler bool
+}
+
+// ExtenderTLSConfig contains settings to enable TLS with extender
+type ExtenderTLSConfig struct {
+	// Server should be accessed without verifying the TLS certificate. For testing only.
+	Insecure bool
+	// ServerName is passed to the server for SNI and is used in the client to check server
+	// certificates against. If ServerName is empty, the hostname used to contact the
+	// server is used.
+	ServerName string
+
+	// Server requires TLS client certificate authentication
+	CertFile string
+	// Server requires TLS client certificate authentication
+	KeyFile string
+	// Trusted root certificates for server
+	CAFile string
+
+	// CertData holds PEM-encoded bytes (typically read from a client certificate file).
+	// CertData takes precedence over CertFile
+	CertData []byte
+	// KeyData holds PEM-encoded bytes (typically read from a client certificate key file).
+	// KeyData takes precedence over KeyFile
+	KeyData []byte
+	// CAData holds PEM-encoded bytes (typically read from a root certificates bundle).
+	// CAData takes precedence over CAFile
+	CAData []byte
+}
+
+// ExtenderConfig holds the parameters used to communicate with the extender. If a verb is unspecified/empty,
+// it is assumed that the extender chose not to provide that extension.
+type ExtenderConfig struct {
+	// URLPrefix at which the extender is available
+	URLPrefix string
+	// Verb for the filter call, empty if not supported. This verb is appended to the URLPrefix when issuing the filter call to extender.
+	FilterVerb string
+	// Verb for the preempt call, empty if not supported. This verb is appended to the URLPrefix when issuing the preempt call to extender.
+	PreemptVerb string
+	// Verb for the prioritize call, empty if not supported. This verb is appended to the URLPrefix when issuing the prioritize call to extender.
+	PrioritizeVerb string
+	// The numeric multiplier for the node scores that the prioritize call generates.
+	// The weight should be a positive integer
+	Weight int
+	// Verb for the bind call, empty if not supported. This verb is appended to the URLPrefix when issuing the bind call to extender.
+	// If this method is implemented by the extender, it is the extender's responsibility to bind the pod to apiserver. Only one extender
+	// can implement this function.
+	BindVerb string
+	// EnableHTTPS specifies whether https should be used to communicate with the extender
+	EnableHTTPS bool
+	// TLSConfig specifies the transport layer security config
+	TLSConfig *ExtenderTLSConfig
+	// HTTPTimeout specifies the timeout duration for a call to the extender. Filter timeout fails the scheduling of the pod. Prioritize
+	// timeout is ignored, k8s/other extenders priorities are used to select the node.
+	HTTPTimeout time.Duration
+	// NodeCacheCapable specifies that the extender is capable of caching node information,
+	// so the scheduler should only send minimal information about the eligible nodes
+	// assuming that the extender already cached full details of all nodes in the cluster
+	NodeCacheCapable bool
+	// ManagedResources is a list of extended resources that are managed by
+	// this extender.
+	// - A pod will be sent to the extender on the Filter, Prioritize and Bind
+	//   (if the extender is the binder) phases iff the pod requests at least
+	//   one of the extended resources in this list. If empty or unspecified,
+	//   all pods will be sent to this extender.
+	// - If IgnoredByScheduler is set to true for a resource, kube-scheduler
+	//   will skip checking the resource in predicates.
+	// +optional
+	ManagedResources []ExtenderManagedResource
+	// Ignorable specifies if the extender is ignorable, i.e. scheduling should not
+	// fail when the extender returns an error or is not reachable.
+	Ignorable bool
+}
+
+// ExtenderPreemptionResult represents the result returned by preemption phase of extender.
+type ExtenderPreemptionResult struct {
+	NodeNameToMetaVictims map[string]*MetaVictims
+}
+
+// ExtenderPreemptionArgs represents the arguments needed by the extender to preempt pods on nodes.
+type ExtenderPreemptionArgs struct {
+	// Pod being scheduled
+	Pod *v1.Pod
+	// Victims map generated by scheduler preemption phase
+	// Only set NodeNameToMetaVictims if ExtenderConfig.NodeCacheCapable == true. Otherwise, only set NodeNameToVictims.
+	NodeNameToVictims     map[string]*Victims
+	NodeNameToMetaVictims map[string]*MetaVictims
+}
+
+// Victims represents:
+//   pods:  a group of pods expected to be preempted.
+//   numPDBViolations: the count of violations of PodDisruptionBudget
+type Victims struct {
+	Pods             []*v1.Pod
+	NumPDBViolations int
+}
+
+// MetaPod represent identifier for a v1.Pod
+type MetaPod struct {
+	UID string
+}
+
+// MetaVictims represents:
+//   pods:  a group of pods expected to be preempted.
+//     Only Pod identifiers will be sent and user are expect to get v1.Pod in their own way.
+//   numPDBViolations: the count of violations of PodDisruptionBudget
+type MetaVictims struct {
+	Pods             []*MetaPod
+	NumPDBViolations int
+}
+
+// ExtenderArgs represents the arguments needed by the extender to filter/prioritize
+// nodes for a pod.
+type ExtenderArgs struct {
+	// Pod being scheduled
+	Pod *v1.Pod
+	// List of candidate nodes where the pod can be scheduled; to be populated
+	// only if ExtenderConfig.NodeCacheCapable == false
+	Nodes *v1.NodeList
+	// List of candidate node names where the pod can be scheduled; to be
+	// populated only if ExtenderConfig.NodeCacheCapable == true
+	NodeNames *[]string
+}
+
+// FailedNodesMap represents the filtered out nodes, with node names and failure messages
+type FailedNodesMap map[string]string
+
+// ExtenderFilterResult represents the results of a filter call to an extender
+type ExtenderFilterResult struct {
+	// Filtered set of nodes where the pod can be scheduled; to be populated
+	// only if ExtenderConfig.NodeCacheCapable == false
+	Nodes *v1.NodeList
+	// Filtered set of nodes where the pod can be scheduled; to be populated
+	// only if ExtenderConfig.NodeCacheCapable == true
+	NodeNames *[]string
+	// Filtered out nodes where the pod can't be scheduled and the failure messages
+	FailedNodes FailedNodesMap
+	// Error message indicating failure
+	Error string
+}
+
+// ExtenderBindingArgs represents the arguments to an extender for binding a pod to a node.
+type ExtenderBindingArgs struct {
+	// PodName is the name of the pod being bound
+	PodName string
+	// PodNamespace is the namespace of the pod being bound
+	PodNamespace string
+	// PodUID is the UID of the pod being bound
+	PodUID types.UID
+	// Node selected by the scheduler
+	Node string
+}
+
+// ExtenderBindingResult represents the result of binding of a pod to a node from an extender.
+type ExtenderBindingResult struct {
+	// Error message indicating failure
+	Error string
+}
+
+// HostPriority represents the priority of scheduling to a particular host, higher priority is better.
+type HostPriority struct {
+	// Name of the host
+	Host string
+	// Score associated with the host
+	Score int
+}
+
+// HostPriorityList declares a []HostPriority type.
+type HostPriorityList []HostPriority
+
+func (h HostPriorityList) Len() int {
+	return len(h)
+}
+
+func (h HostPriorityList) Less(i, j int) bool {
+	if h[i].Score == h[j].Score {
+		return h[i].Host < h[j].Host
+	}
+	return h[i].Score < h[j].Score
+}
+
+func (h HostPriorityList) Swap(i, j int) {
+	h[i], h[j] = h[j], h[i]
+}

vendor/k8s.io/kubernetes/pkg/scheduler/api/well_known_labels.go

@@ -0,0 +1,72 @@
+/*
+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 api
+
+import (
+	api "k8s.io/kubernetes/pkg/apis/core"
+)
+
+const (
+	// TaintNodeNotReady will be added when node is not ready
+	// and feature-gate for TaintBasedEvictions flag is enabled,
+	// and removed when node becomes ready.
+	TaintNodeNotReady = "node.kubernetes.io/not-ready"
+
+	// TaintNodeUnreachable will be added when node becomes unreachable
+	// (corresponding to NodeReady status ConditionUnknown)
+	// and feature-gate for TaintBasedEvictions flag is enabled,
+	// and removed when node becomes reachable (NodeReady status ConditionTrue).
+	TaintNodeUnreachable = "node.kubernetes.io/unreachable"
+
+	// TaintNodeUnschedulable will be added when node becomes unschedulable
+	// and feature-gate for TaintNodesByCondition flag is enabled,
+	// and removed when node becomes scheduable.
+	TaintNodeUnschedulable = "node.kubernetes.io/unschedulable"
+
+	// TaintNodeMemoryPressure will be added when node has memory pressure
+	// and feature-gate for TaintNodesByCondition flag is enabled,
+	// and removed when node has enough memory.
+	TaintNodeMemoryPressure = "node.kubernetes.io/memory-pressure"
+
+	// TaintNodeDiskPressure will be added when node has disk pressure
+	// and feature-gate for TaintNodesByCondition flag is enabled,
+	// and removed when node has enough disk.
+	TaintNodeDiskPressure = "node.kubernetes.io/disk-pressure"
+
+	// TaintNodeNetworkUnavailable will be added when node's network is unavailable
+	// and feature-gate for TaintNodesByCondition flag is enabled,
+	// and removed when network becomes ready.
+	TaintNodeNetworkUnavailable = "node.kubernetes.io/network-unavailable"
+
+	// TaintNodePIDPressure will be added when node has pid pressure
+	// and feature-gate for TaintNodesByCondition flag is enabled,
+	// and removed when node has enough disk.
+	TaintNodePIDPressure = "node.kubernetes.io/pid-pressure"
+
+	// TaintExternalCloudProvider sets this taint on a node to mark it as unusable,
+	// when kubelet is started with the "external" cloud provider, until a controller
+	// from the cloud-controller-manager intitializes this node, and then removes
+	// the taint
+	TaintExternalCloudProvider = "node.cloudprovider.kubernetes.io/uninitialized"
+
+	// TaintNodeShutdown when node is shutdown in external cloud provider
+	TaintNodeShutdown = "node.cloudprovider.kubernetes.io/shutdown"
+
+	// NodeFieldSelectorKeyNodeName ('metadata.name') uses this as node field selector key
+	// when selecting node by node's name.
+	NodeFieldSelectorKeyNodeName = api.ObjectNameField
+)

vendor/k8s.io/kubernetes/pkg/scheduler/api/zz_generated.deepcopy.go

@@ -0,0 +1,669 @@
+// +build !ignore_autogenerated
+
+/*
+Copyright 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.
+*/
+
+// Code generated by deepcopy-gen. DO NOT EDIT.
+
+package api
+
+import (
+	v1 "k8s.io/api/core/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 *ExtenderArgs) DeepCopyInto(out *ExtenderArgs) {
+	*out = *in
+	if in.Pod != nil {
+		in, out := &in.Pod, &out.Pod
+		*out = new(v1.Pod)
+		(*in).DeepCopyInto(*out)
+	}
+	if in.Nodes != nil {
+		in, out := &in.Nodes, &out.Nodes
+		*out = new(v1.NodeList)
+		(*in).DeepCopyInto(*out)
+	}
+	if in.NodeNames != nil {
+		in, out := &in.NodeNames, &out.NodeNames
+		*out = new([]string)
+		if **in != nil {
+			in, out := *in, *out
+			*out = make([]string, len(*in))
+			copy(*out, *in)
+		}
+	}
+	return
+}
+
+// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new ExtenderArgs.
+func (in *ExtenderArgs) DeepCopy() *ExtenderArgs {
+	if in == nil {
+		return nil
+	}
+	out := new(ExtenderArgs)
+	in.DeepCopyInto(out)
+	return out
+}
+
+// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
+func (in *ExtenderBindingArgs) DeepCopyInto(out *ExtenderBindingArgs) {
+	*out = *in
+	return
+}
+
+// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new ExtenderBindingArgs.
+func (in *ExtenderBindingArgs) DeepCopy() *ExtenderBindingArgs {
+	if in == nil {
+		return nil
+	}
+	out := new(ExtenderBindingArgs)
+	in.DeepCopyInto(out)
+	return out
+}
+
+// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
+func (in *ExtenderBindingResult) DeepCopyInto(out *ExtenderBindingResult) {
+	*out = *in
+	return
+}
+
+// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new ExtenderBindingResult.
+func (in *ExtenderBindingResult) DeepCopy() *ExtenderBindingResult {
+	if in == nil {
+		return nil
+	}
+	out := new(ExtenderBindingResult)
+	in.DeepCopyInto(out)
+	return out
+}
+
+// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
+func (in *ExtenderConfig) DeepCopyInto(out *ExtenderConfig) {
+	*out = *in
+	if in.TLSConfig != nil {
+		in, out := &in.TLSConfig, &out.TLSConfig
+		*out = new(ExtenderTLSConfig)
+		(*in).DeepCopyInto(*out)
+	}
+	if in.ManagedResources != nil {
+		in, out := &in.ManagedResources, &out.ManagedResources
+		*out = make([]ExtenderManagedResource, len(*in))
+		copy(*out, *in)
+	}
+	return
+}
+
+// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new ExtenderConfig.
+func (in *ExtenderConfig) DeepCopy() *ExtenderConfig {
+	if in == nil {
+		return nil
+	}
+	out := new(ExtenderConfig)
+	in.DeepCopyInto(out)
+	return out
+}
+
+// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
+func (in *ExtenderFilterResult) DeepCopyInto(out *ExtenderFilterResult) {
+	*out = *in
+	if in.Nodes != nil {
+		in, out := &in.Nodes, &out.Nodes
+		*out = new(v1.NodeList)
+		(*in).DeepCopyInto(*out)
+	}
+	if in.NodeNames != nil {
+		in, out := &in.NodeNames, &out.NodeNames
+		*out = new([]string)
+		if **in != nil {
+			in, out := *in, *out
+			*out = make([]string, len(*in))
+			copy(*out, *in)
+		}
+	}
+	if in.FailedNodes != nil {
+		in, out := &in.FailedNodes, &out.FailedNodes
+		*out = make(FailedNodesMap, len(*in))
+		for key, val := range *in {
+			(*out)[key] = val
+		}
+	}
+	return
+}
+
+// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new ExtenderFilterResult.
+func (in *ExtenderFilterResult) DeepCopy() *ExtenderFilterResult {
+	if in == nil {
+		return nil
+	}
+	out := new(ExtenderFilterResult)
+	in.DeepCopyInto(out)
+	return out
+}
+
+// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
+func (in *ExtenderManagedResource) DeepCopyInto(out *ExtenderManagedResource) {
+	*out = *in
+	return
+}
+
+// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new ExtenderManagedResource.
+func (in *ExtenderManagedResource) DeepCopy() *ExtenderManagedResource {
+	if in == nil {
+		return nil
+	}
+	out := new(ExtenderManagedResource)
+	in.DeepCopyInto(out)
+	return out
+}
+
+// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
+func (in *ExtenderPreemptionArgs) DeepCopyInto(out *ExtenderPreemptionArgs) {
+	*out = *in
+	if in.Pod != nil {
+		in, out := &in.Pod, &out.Pod
+		*out = new(v1.Pod)
+		(*in).DeepCopyInto(*out)
+	}
+	if in.NodeNameToVictims != nil {
+		in, out := &in.NodeNameToVictims, &out.NodeNameToVictims
+		*out = make(map[string]*Victims, len(*in))
+		for key, val := range *in {
+			var outVal *Victims
+			if val == nil {
+				(*out)[key] = nil
+			} else {
+				in, out := &val, &outVal
+				*out = new(Victims)
+				(*in).DeepCopyInto(*out)
+			}
+			(*out)[key] = outVal
+		}
+	}
+	if in.NodeNameToMetaVictims != nil {
+		in, out := &in.NodeNameToMetaVictims, &out.NodeNameToMetaVictims
+		*out = make(map[string]*MetaVictims, len(*in))
+		for key, val := range *in {
+			var outVal *MetaVictims
+			if val == nil {
+				(*out)[key] = nil
+			} else {
+				in, out := &val, &outVal
+				*out = new(MetaVictims)
+				(*in).DeepCopyInto(*out)
+			}
+			(*out)[key] = outVal
+		}
+	}
+	return
+}
+
+// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new ExtenderPreemptionArgs.
+func (in *ExtenderPreemptionArgs) DeepCopy() *ExtenderPreemptionArgs {
+	if in == nil {
+		return nil
+	}
+	out := new(ExtenderPreemptionArgs)
+	in.DeepCopyInto(out)
+	return out
+}
+
+// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
+func (in *ExtenderPreemptionResult) DeepCopyInto(out *ExtenderPreemptionResult) {
+	*out = *in
+	if in.NodeNameToMetaVictims != nil {
+		in, out := &in.NodeNameToMetaVictims, &out.NodeNameToMetaVictims
+		*out = make(map[string]*MetaVictims, len(*in))
+		for key, val := range *in {
+			var outVal *MetaVictims
+			if val == nil {
+				(*out)[key] = nil
+			} else {
+				in, out := &val, &outVal
+				*out = new(MetaVictims)
+				(*in).DeepCopyInto(*out)
+			}
+			(*out)[key] = outVal
+		}
+	}
+	return
+}
+
+// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new ExtenderPreemptionResult.
+func (in *ExtenderPreemptionResult) DeepCopy() *ExtenderPreemptionResult {
+	if in == nil {
+		return nil
+	}
+	out := new(ExtenderPreemptionResult)
+	in.DeepCopyInto(out)
+	return out
+}
+
+// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
+func (in *ExtenderTLSConfig) DeepCopyInto(out *ExtenderTLSConfig) {
+	*out = *in
+	if in.CertData != nil {
+		in, out := &in.CertData, &out.CertData
+		*out = make([]byte, len(*in))
+		copy(*out, *in)
+	}
+	if in.KeyData != nil {
+		in, out := &in.KeyData, &out.KeyData
+		*out = make([]byte, len(*in))
+		copy(*out, *in)
+	}
+	if in.CAData != nil {
+		in, out := &in.CAData, &out.CAData
+		*out = make([]byte, len(*in))
+		copy(*out, *in)
+	}
+	return
+}
+
+// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new ExtenderTLSConfig.
+func (in *ExtenderTLSConfig) DeepCopy() *ExtenderTLSConfig {
+	if in == nil {
+		return nil
+	}
+	out := new(ExtenderTLSConfig)
+	in.DeepCopyInto(out)
+	return out
+}
+
+// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
+func (in FailedNodesMap) DeepCopyInto(out *FailedNodesMap) {
+	{
+		in := &in
+		*out = make(FailedNodesMap, len(*in))
+		for key, val := range *in {
+			(*out)[key] = val
+		}
+		return
+	}
+}
+
+// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new FailedNodesMap.
+func (in FailedNodesMap) DeepCopy() FailedNodesMap {
+	if in == nil {
+		return nil
+	}
+	out := new(FailedNodesMap)
+	in.DeepCopyInto(out)
+	return *out
+}
+
+// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
+func (in *HostPriority) DeepCopyInto(out *HostPriority) {
+	*out = *in
+	return
+}
+
+// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new HostPriority.
+func (in *HostPriority) DeepCopy() *HostPriority {
+	if in == nil {
+		return nil
+	}
+	out := new(HostPriority)
+	in.DeepCopyInto(out)
+	return out
+}
+
+// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
+func (in HostPriorityList) DeepCopyInto(out *HostPriorityList) {
+	{
+		in := &in
+		*out = make(HostPriorityList, len(*in))
+		copy(*out, *in)
+		return
+	}
+}
+
+// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new HostPriorityList.
+func (in HostPriorityList) DeepCopy() HostPriorityList {
+	if in == nil {
+		return nil
+	}
+	out := new(HostPriorityList)
+	in.DeepCopyInto(out)
+	return *out
+}
+
+// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
+func (in *LabelPreference) DeepCopyInto(out *LabelPreference) {
+	*out = *in
+	return
+}
+
+// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new LabelPreference.
+func (in *LabelPreference) DeepCopy() *LabelPreference {
+	if in == nil {
+		return nil
+	}
+	out := new(LabelPreference)
+	in.DeepCopyInto(out)
+	return out
+}
+
+// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
+func (in *LabelsPresence) DeepCopyInto(out *LabelsPresence) {
+	*out = *in
+	if in.Labels != nil {
+		in, out := &in.Labels, &out.Labels
+		*out = make([]string, len(*in))
+		copy(*out, *in)
+	}
+	return
+}
+
+// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new LabelsPresence.
+func (in *LabelsPresence) DeepCopy() *LabelsPresence {
+	if in == nil {
+		return nil
+	}
+	out := new(LabelsPresence)
+	in.DeepCopyInto(out)
+	return out
+}
+
+// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
+func (in *MetaPod) DeepCopyInto(out *MetaPod) {
+	*out = *in
+	return
+}
+
+// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new MetaPod.
+func (in *MetaPod) DeepCopy() *MetaPod {
+	if in == nil {
+		return nil
+	}
+	out := new(MetaPod)
+	in.DeepCopyInto(out)
+	return out
+}
+
+// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
+func (in *MetaVictims) DeepCopyInto(out *MetaVictims) {
+	*out = *in
+	if in.Pods != nil {
+		in, out := &in.Pods, &out.Pods
+		*out = make([]*MetaPod, len(*in))
+		for i := range *in {
+			if (*in)[i] != nil {
+				in, out := &(*in)[i], &(*out)[i]
+				*out = new(MetaPod)
+				**out = **in
+			}
+		}
+	}
+	return
+}
+
+// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new MetaVictims.
+func (in *MetaVictims) DeepCopy() *MetaVictims {
+	if in == nil {
+		return nil
+	}
+	out := new(MetaVictims)
+	in.DeepCopyInto(out)
+	return out
+}
+
+// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
+func (in *Policy) DeepCopyInto(out *Policy) {
+	*out = *in
+	out.TypeMeta = in.TypeMeta
+	if in.Predicates != nil {
+		in, out := &in.Predicates, &out.Predicates
+		*out = make([]PredicatePolicy, len(*in))
+		for i := range *in {
+			(*in)[i].DeepCopyInto(&(*out)[i])
+		}
+	}
+	if in.Priorities != nil {
+		in, out := &in.Priorities, &out.Priorities
+		*out = make([]PriorityPolicy, len(*in))
+		for i := range *in {
+			(*in)[i].DeepCopyInto(&(*out)[i])
+		}
+	}
+	if in.ExtenderConfigs != nil {
+		in, out := &in.ExtenderConfigs, &out.ExtenderConfigs
+		*out = make([]ExtenderConfig, len(*in))
+		for i := range *in {
+			(*in)[i].DeepCopyInto(&(*out)[i])
+		}
+	}
+	return
+}
+
+// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new Policy.
+func (in *Policy) DeepCopy() *Policy {
+	if in == nil {
+		return nil
+	}
+	out := new(Policy)
+	in.DeepCopyInto(out)
+	return out
+}
+
+// DeepCopyObject is an autogenerated deepcopy function, copying the receiver, creating a new runtime.Object.
+func (in *Policy) 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 *PredicateArgument) DeepCopyInto(out *PredicateArgument) {
+	*out = *in
+	if in.ServiceAffinity != nil {
+		in, out := &in.ServiceAffinity, &out.ServiceAffinity
+		*out = new(ServiceAffinity)
+		(*in).DeepCopyInto(*out)
+	}
+	if in.LabelsPresence != nil {
+		in, out := &in.LabelsPresence, &out.LabelsPresence
+		*out = new(LabelsPresence)
+		(*in).DeepCopyInto(*out)
+	}
+	return
+}
+
+// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new PredicateArgument.
+func (in *PredicateArgument) DeepCopy() *PredicateArgument {
+	if in == nil {
+		return nil
+	}
+	out := new(PredicateArgument)
+	in.DeepCopyInto(out)
+	return out
+}
+
+// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
+func (in *PredicatePolicy) DeepCopyInto(out *PredicatePolicy) {
+	*out = *in
+	if in.Argument != nil {
+		in, out := &in.Argument, &out.Argument
+		*out = new(PredicateArgument)
+		(*in).DeepCopyInto(*out)
+	}
+	return
+}
+
+// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new PredicatePolicy.
+func (in *PredicatePolicy) DeepCopy() *PredicatePolicy {
+	if in == nil {
+		return nil
+	}
+	out := new(PredicatePolicy)
+	in.DeepCopyInto(out)
+	return out
+}
+
+// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
+func (in *PriorityArgument) DeepCopyInto(out *PriorityArgument) {
+	*out = *in
+	if in.ServiceAntiAffinity != nil {
+		in, out := &in.ServiceAntiAffinity, &out.ServiceAntiAffinity
+		*out = new(ServiceAntiAffinity)
+		**out = **in
+	}
+	if in.LabelPreference != nil {
+		in, out := &in.LabelPreference, &out.LabelPreference
+		*out = new(LabelPreference)
+		**out = **in
+	}
+	if in.RequestedToCapacityRatioArguments != nil {
+		in, out := &in.RequestedToCapacityRatioArguments, &out.RequestedToCapacityRatioArguments
+		*out = new(RequestedToCapacityRatioArguments)
+		(*in).DeepCopyInto(*out)
+	}
+	return
+}
+
+// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new PriorityArgument.
+func (in *PriorityArgument) DeepCopy() *PriorityArgument {
+	if in == nil {
+		return nil
+	}
+	out := new(PriorityArgument)
+	in.DeepCopyInto(out)
+	return out
+}
+
+// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
+func (in *PriorityPolicy) DeepCopyInto(out *PriorityPolicy) {
+	*out = *in
+	if in.Argument != nil {
+		in, out := &in.Argument, &out.Argument
+		*out = new(PriorityArgument)
+		(*in).DeepCopyInto(*out)
+	}
+	return
+}
+
+// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new PriorityPolicy.
+func (in *PriorityPolicy) DeepCopy() *PriorityPolicy {
+	if in == nil {
+		return nil
+	}
+	out := new(PriorityPolicy)
+	in.DeepCopyInto(out)
+	return out
+}
+
+// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
+func (in *RequestedToCapacityRatioArguments) DeepCopyInto(out *RequestedToCapacityRatioArguments) {
+	*out = *in
+	if in.UtilizationShape != nil {
+		in, out := &in.UtilizationShape, &out.UtilizationShape
+		*out = make([]UtilizationShapePoint, len(*in))
+		copy(*out, *in)
+	}
+	return
+}
+
+// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new RequestedToCapacityRatioArguments.
+func (in *RequestedToCapacityRatioArguments) DeepCopy() *RequestedToCapacityRatioArguments {
+	if in == nil {
+		return nil
+	}
+	out := new(RequestedToCapacityRatioArguments)
+	in.DeepCopyInto(out)
+	return out
+}
+
+// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
+func (in *ServiceAffinity) DeepCopyInto(out *ServiceAffinity) {
+	*out = *in
+	if in.Labels != nil {
+		in, out := &in.Labels, &out.Labels
+		*out = make([]string, len(*in))
+		copy(*out, *in)
+	}
+	return
+}
+
+// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new ServiceAffinity.
+func (in *ServiceAffinity) DeepCopy() *ServiceAffinity {
+	if in == nil {
+		return nil
+	}
+	out := new(ServiceAffinity)
+	in.DeepCopyInto(out)
+	return out
+}
+
+// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
+func (in *ServiceAntiAffinity) DeepCopyInto(out *ServiceAntiAffinity) {
+	*out = *in
+	return
+}
+
+// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new ServiceAntiAffinity.
+func (in *ServiceAntiAffinity) DeepCopy() *ServiceAntiAffinity {
+	if in == nil {
+		return nil
+	}
+	out := new(ServiceAntiAffinity)
+	in.DeepCopyInto(out)
+	return out
+}
+
+// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
+func (in *UtilizationShapePoint) DeepCopyInto(out *UtilizationShapePoint) {
+	*out = *in
+	return
+}
+
+// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new UtilizationShapePoint.
+func (in *UtilizationShapePoint) DeepCopy() *UtilizationShapePoint {
+	if in == nil {
+		return nil
+	}
+	out := new(UtilizationShapePoint)
+	in.DeepCopyInto(out)
+	return out
+}
+
+// DeepCopyInto is an autogenerated deepcopy function, copying the receiver, writing into out. in must be non-nil.
+func (in *Victims) DeepCopyInto(out *Victims) {
+	*out = *in
+	if in.Pods != nil {
+		in, out := &in.Pods, &out.Pods
+		*out = make([]*v1.Pod, len(*in))
+		for i := range *in {
+			if (*in)[i] != nil {
+				in, out := &(*in)[i], &(*out)[i]
+				*out = new(v1.Pod)
+				(*in).DeepCopyInto(*out)
+			}
+		}
+	}
+	return
+}
+
+// DeepCopy is an autogenerated deepcopy function, copying the receiver, creating a new Victims.
+func (in *Victims) DeepCopy() *Victims {
+	if in == nil {
+		return nil
+	}
+	out := new(Victims)
+	in.DeepCopyInto(out)
+	return out
+}

vendor/k8s.io/kubernetes/pkg/scheduler/metrics/metrics.go

@@ -0,0 +1,244 @@
+/*
+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 metrics
+
+import (
+	"sync"
+	"time"
+
+	"github.com/prometheus/client_golang/prometheus"
+	"k8s.io/kubernetes/pkg/controller/volume/persistentvolume"
+)
+
+const (
+	// SchedulerSubsystem - subsystem name used by scheduler
+	SchedulerSubsystem = "scheduler"
+	// SchedulingLatencyName - scheduler latency metric name
+	SchedulingLatencyName = "scheduling_duration_seconds"
+	// DeprecatedSchedulingLatencyName - scheduler latency metric name which is deprecated
+	DeprecatedSchedulingLatencyName = "scheduling_latency_seconds"
+
+	// OperationLabel - operation label name
+	OperationLabel = "operation"
+	// Below are possible values for the operation label. Each represents a substep of e2e scheduling:
+
+	// PredicateEvaluation - predicate evaluation operation label value
+	PredicateEvaluation = "predicate_evaluation"
+	// PriorityEvaluation - priority evaluation operation label value
+	PriorityEvaluation = "priority_evaluation"
+	// PreemptionEvaluation - preemption evaluation operation label value (occurs in case of scheduling fitError).
+	PreemptionEvaluation = "preemption_evaluation"
+	// Binding - binding operation label value
+	Binding = "binding"
+	// E2eScheduling - e2e scheduling operation label value
+)
+
+// All the histogram based metrics have 1ms as size for the smallest bucket.
+var (
+	scheduleAttempts = prometheus.NewCounterVec(
+		prometheus.CounterOpts{
+			Subsystem: SchedulerSubsystem,
+			Name:      "schedule_attempts_total",
+			Help:      "Number of attempts to schedule pods, by the result. 'unschedulable' means a pod could not be scheduled, while 'error' means an internal scheduler problem.",
+		}, []string{"result"})
+	// PodScheduleSuccesses counts how many pods were scheduled.
+	PodScheduleSuccesses = scheduleAttempts.With(prometheus.Labels{"result": "scheduled"})
+	// PodScheduleFailures counts how many pods could not be scheduled.
+	PodScheduleFailures = scheduleAttempts.With(prometheus.Labels{"result": "unschedulable"})
+	// PodScheduleErrors counts how many pods could not be scheduled due to a scheduler error.
+	PodScheduleErrors = scheduleAttempts.With(prometheus.Labels{"result": "error"})
+	SchedulingLatency = prometheus.NewSummaryVec(
+		prometheus.SummaryOpts{
+			Subsystem: SchedulerSubsystem,
+			Name:      SchedulingLatencyName,
+			Help:      "Scheduling latency in seconds split by sub-parts of the scheduling operation",
+			// Make the sliding window of 5h.
+			// TODO: The value for this should be based on some SLI definition (long term).
+			MaxAge: 5 * time.Hour,
+		},
+		[]string{OperationLabel},
+	)
+	DeprecatedSchedulingLatency = prometheus.NewSummaryVec(
+		prometheus.SummaryOpts{
+			Subsystem: SchedulerSubsystem,
+			Name:      DeprecatedSchedulingLatencyName,
+			Help:      "(Deprecated) Scheduling latency in seconds split by sub-parts of the scheduling operation",
+			// Make the sliding window of 5h.
+			// TODO: The value for this should be based on some SLI definition (long term).
+			MaxAge: 5 * time.Hour,
+		},
+		[]string{OperationLabel},
+	)
+	E2eSchedulingLatency = prometheus.NewHistogram(
+		prometheus.HistogramOpts{
+			Subsystem: SchedulerSubsystem,
+			Name:      "e2e_scheduling_duration_seconds",
+			Help:      "E2e scheduling latency in seconds (scheduling algorithm + binding)",
+			Buckets:   prometheus.ExponentialBuckets(0.001, 2, 15),
+		},
+	)
+	DeprecatedE2eSchedulingLatency = prometheus.NewHistogram(
+		prometheus.HistogramOpts{
+			Subsystem: SchedulerSubsystem,
+			Name:      "e2e_scheduling_latency_microseconds",
+			Help:      "(Deprecated) E2e scheduling latency in microseconds (scheduling algorithm + binding)",
+			Buckets:   prometheus.ExponentialBuckets(1000, 2, 15),
+		},
+	)
+	SchedulingAlgorithmLatency = prometheus.NewHistogram(
+		prometheus.HistogramOpts{
+			Subsystem: SchedulerSubsystem,
+			Name:      "scheduling_algorithm_duration_seconds",
+			Help:      "Scheduling algorithm latency in seconds",
+			Buckets:   prometheus.ExponentialBuckets(0.001, 2, 15),
+		},
+	)
+	DeprecatedSchedulingAlgorithmLatency = prometheus.NewHistogram(
+		prometheus.HistogramOpts{
+			Subsystem: SchedulerSubsystem,
+			Name:      "scheduling_algorithm_latency_microseconds",
+			Help:      "(Deprecated) Scheduling algorithm latency in microseconds",
+			Buckets:   prometheus.ExponentialBuckets(1000, 2, 15),
+		},
+	)
+	SchedulingAlgorithmPredicateEvaluationDuration = prometheus.NewHistogram(
+		prometheus.HistogramOpts{
+			Subsystem: SchedulerSubsystem,
+			Name:      "scheduling_algorithm_predicate_evaluation_seconds",
+			Help:      "Scheduling algorithm predicate evaluation duration in seconds",
+			Buckets:   prometheus.ExponentialBuckets(0.001, 2, 15),
+		},
+	)
+	DeprecatedSchedulingAlgorithmPredicateEvaluationDuration = prometheus.NewHistogram(
+		prometheus.HistogramOpts{
+			Subsystem: SchedulerSubsystem,
+			Name:      "scheduling_algorithm_predicate_evaluation",
+			Help:      "(Deprecated) Scheduling algorithm predicate evaluation duration in microseconds",
+			Buckets:   prometheus.ExponentialBuckets(1000, 2, 15),
+		},
+	)
+	SchedulingAlgorithmPriorityEvaluationDuration = prometheus.NewHistogram(
+		prometheus.HistogramOpts{
+			Subsystem: SchedulerSubsystem,
+			Name:      "scheduling_algorithm_priority_evaluation_seconds",
+			Help:      "Scheduling algorithm priority evaluation duration in seconds",
+			Buckets:   prometheus.ExponentialBuckets(0.001, 2, 15),
+		},
+	)
+	DeprecatedSchedulingAlgorithmPriorityEvaluationDuration = prometheus.NewHistogram(
+		prometheus.HistogramOpts{
+			Subsystem: SchedulerSubsystem,
+			Name:      "scheduling_algorithm_priority_evaluation",
+			Help:      "(Deprecated) Scheduling algorithm priority evaluation duration in microseconds",
+			Buckets:   prometheus.ExponentialBuckets(1000, 2, 15),
+		},
+	)
+	SchedulingAlgorithmPremptionEvaluationDuration = prometheus.NewHistogram(
+		prometheus.HistogramOpts{
+			Subsystem: SchedulerSubsystem,
+			Name:      "scheduling_algorithm_preemption_evaluation_seconds",
+			Help:      "Scheduling algorithm preemption evaluation duration in seconds",
+			Buckets:   prometheus.ExponentialBuckets(0.001, 2, 15),
+		},
+	)
+	DeprecatedSchedulingAlgorithmPremptionEvaluationDuration = prometheus.NewHistogram(
+		prometheus.HistogramOpts{
+			Subsystem: SchedulerSubsystem,
+			Name:      "scheduling_algorithm_preemption_evaluation",
+			Help:      "(Deprecated) Scheduling algorithm preemption evaluation duration in microseconds",
+			Buckets:   prometheus.ExponentialBuckets(1000, 2, 15),
+		},
+	)
+	BindingLatency = prometheus.NewHistogram(
+		prometheus.HistogramOpts{
+			Subsystem: SchedulerSubsystem,
+			Name:      "binding_duration_seconds",
+			Help:      "Binding latency in seconds",
+			Buckets:   prometheus.ExponentialBuckets(0.001, 2, 15),
+		},
+	)
+	DeprecatedBindingLatency = prometheus.NewHistogram(
+		prometheus.HistogramOpts{
+			Subsystem: SchedulerSubsystem,
+			Name:      "binding_latency_microseconds",
+			Help:      "(Deprecated) Binding latency in microseconds",
+			Buckets:   prometheus.ExponentialBuckets(1000, 2, 15),
+		},
+	)
+	PreemptionVictims = prometheus.NewGauge(
+		prometheus.GaugeOpts{
+			Subsystem: SchedulerSubsystem,
+			Name:      "pod_preemption_victims",
+			Help:      "Number of selected preemption victims",
+		})
+	PreemptionAttempts = prometheus.NewCounter(
+		prometheus.CounterOpts{
+			Subsystem: SchedulerSubsystem,
+			Name:      "total_preemption_attempts",
+			Help:      "Total preemption attempts in the cluster till now",
+		})
+
+	metricsList = []prometheus.Collector{
+		scheduleAttempts,
+		SchedulingLatency,
+		DeprecatedSchedulingLatency,
+		E2eSchedulingLatency,
+		DeprecatedE2eSchedulingLatency,
+		SchedulingAlgorithmLatency,
+		DeprecatedSchedulingAlgorithmLatency,
+		BindingLatency,
+		DeprecatedBindingLatency,
+		SchedulingAlgorithmPredicateEvaluationDuration,
+		DeprecatedSchedulingAlgorithmPredicateEvaluationDuration,
+		SchedulingAlgorithmPriorityEvaluationDuration,
+		DeprecatedSchedulingAlgorithmPriorityEvaluationDuration,
+		SchedulingAlgorithmPremptionEvaluationDuration,
+		DeprecatedSchedulingAlgorithmPremptionEvaluationDuration,
+		PreemptionVictims,
+		PreemptionAttempts,
+	}
+)
+
+var registerMetrics sync.Once
+
+// Register all metrics.
+func Register() {
+	// Register the metrics.
+	registerMetrics.Do(func() {
+		for _, metric := range metricsList {
+			prometheus.MustRegister(metric)
+		}
+
+		persistentvolume.RegisterVolumeSchedulingMetrics()
+	})
+}
+
+// Reset resets metrics
+func Reset() {
+	SchedulingLatency.Reset()
+	DeprecatedSchedulingLatency.Reset()
+}
+
+// SinceInMicroseconds gets the time since the specified start in microseconds.
+func SinceInMicroseconds(start time.Time) float64 {
+	return float64(time.Since(start).Nanoseconds() / time.Microsecond.Nanoseconds())
+}
+
+// SinceInSeconds gets the time since the specified start in seconds.
+func SinceInSeconds(start time.Time) float64 {
+	return time.Since(start).Seconds()
+}

vendor/k8s.io/kubernetes/pkg/scheduler/nodeinfo/host_ports.go

@@ -0,0 +1,135 @@
+/*
+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 nodeinfo
+
+import (
+	"k8s.io/api/core/v1"
+)
+
+// DefaultBindAllHostIP defines the default ip address used to bind to all host.
+const DefaultBindAllHostIP = "0.0.0.0"
+
+// ProtocolPort represents a protocol port pair, e.g. tcp:80.
+type ProtocolPort struct {
+	Protocol string
+	Port     int32
+}
+
+// NewProtocolPort creates a ProtocolPort instance.
+func NewProtocolPort(protocol string, port int32) *ProtocolPort {
+	pp := &ProtocolPort{
+		Protocol: protocol,
+		Port:     port,
+	}
+
+	if len(pp.Protocol) == 0 {
+		pp.Protocol = string(v1.ProtocolTCP)
+	}
+
+	return pp
+}
+
+// HostPortInfo stores mapping from ip to a set of ProtocolPort
+type HostPortInfo map[string]map[ProtocolPort]struct{}
+
+// Add adds (ip, protocol, port) to HostPortInfo
+func (h HostPortInfo) Add(ip, protocol string, port int32) {
+	if port <= 0 {
+		return
+	}
+
+	h.sanitize(&ip, &protocol)
+
+	pp := NewProtocolPort(protocol, port)
+	if _, ok := h[ip]; !ok {
+		h[ip] = map[ProtocolPort]struct{}{
+			*pp: {},
+		}
+		return
+	}
+
+	h[ip][*pp] = struct{}{}
+}
+
+// Remove removes (ip, protocol, port) from HostPortInfo
+func (h HostPortInfo) Remove(ip, protocol string, port int32) {
+	if port <= 0 {
+		return
+	}
+
+	h.sanitize(&ip, &protocol)
+
+	pp := NewProtocolPort(protocol, port)
+	if m, ok := h[ip]; ok {
+		delete(m, *pp)
+		if len(h[ip]) == 0 {
+			delete(h, ip)
+		}
+	}
+}
+
+// Len returns the total number of (ip, protocol, port) tuple in HostPortInfo
+func (h HostPortInfo) Len() int {
+	length := 0
+	for _, m := range h {
+		length += len(m)
+	}
+	return length
+}
+
+// CheckConflict checks if the input (ip, protocol, port) conflicts with the existing
+// ones in HostPortInfo.
+func (h HostPortInfo) CheckConflict(ip, protocol string, port int32) bool {
+	if port <= 0 {
+		return false
+	}
+
+	h.sanitize(&ip, &protocol)
+
+	pp := NewProtocolPort(protocol, port)
+
+	// If ip is 0.0.0.0 check all IP's (protocol, port) pair
+	if ip == DefaultBindAllHostIP {
+		for _, m := range h {
+			if _, ok := m[*pp]; ok {
+				return true
+			}
+		}
+		return false
+	}
+
+	// If ip isn't 0.0.0.0, only check IP and 0.0.0.0's (protocol, port) pair
+	for _, key := range []string{DefaultBindAllHostIP, ip} {
+		if m, ok := h[key]; ok {
+			if _, ok2 := m[*pp]; ok2 {
+				return true
+			}
+		}
+	}
+
+	return false
+}
+
+// sanitize the parameters
+func (h HostPortInfo) sanitize(ip, protocol *string) {
+	if len(*ip) == 0 {
+		*ip = DefaultBindAllHostIP
+	}
+	if len(*protocol) == 0 {
+		*protocol = string(v1.ProtocolTCP)
+	}
+}

vendor/k8s.io/kubernetes/pkg/scheduler/nodeinfo/node_info.go

@@ -0,0 +1,702 @@
+/*
+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 nodeinfo
+
+import (
+	"errors"
+	"fmt"
+	"sync"
+	"sync/atomic"
+
+	"k8s.io/klog"
+
+	"k8s.io/api/core/v1"
+	"k8s.io/apimachinery/pkg/api/resource"
+	v1helper "k8s.io/kubernetes/pkg/apis/core/v1/helper"
+	priorityutil "k8s.io/kubernetes/pkg/scheduler/algorithm/priorities/util"
+)
+
+var (
+	emptyResource = Resource{}
+	generation    int64
+)
+
+// ImageStateSummary provides summarized information about the state of an image.
+type ImageStateSummary struct {
+	// Size of the image
+	Size int64
+	// Used to track how many nodes have this image
+	NumNodes int
+}
+
+// NodeInfo is node level aggregated information.
+type NodeInfo struct {
+	// Overall node information.
+	node *v1.Node
+
+	pods             []*v1.Pod
+	podsWithAffinity []*v1.Pod
+	usedPorts        HostPortInfo
+
+	// Total requested resource of all pods on this node.
+	// It includes assumed pods which scheduler sends binding to apiserver but
+	// didn't get it as scheduled yet.
+	requestedResource *Resource
+	nonzeroRequest    *Resource
+	// We store allocatedResources (which is Node.Status.Allocatable.*) explicitly
+	// as int64, to avoid conversions and accessing map.
+	allocatableResource *Resource
+
+	// Cached taints of the node for faster lookup.
+	taints    []v1.Taint
+	taintsErr error
+
+	// imageStates holds the entry of an image if and only if this image is on the node. The entry can be used for
+	// checking an image's existence and advanced usage (e.g., image locality scheduling policy) based on the image
+	// state information.
+	imageStates map[string]*ImageStateSummary
+
+	// TransientInfo holds the information pertaining to a scheduling cycle. This will be destructed at the end of
+	// scheduling cycle.
+	// TODO: @ravig. Remove this once we have a clear approach for message passing across predicates and priorities.
+	TransientInfo *TransientSchedulerInfo
+
+	// Cached conditions of node for faster lookup.
+	memoryPressureCondition v1.ConditionStatus
+	diskPressureCondition   v1.ConditionStatus
+	pidPressureCondition    v1.ConditionStatus
+
+	// Whenever NodeInfo changes, generation is bumped.
+	// This is used to avoid cloning it if the object didn't change.
+	generation int64
+}
+
+//initializeNodeTransientInfo initializes transient information pertaining to node.
+func initializeNodeTransientInfo() nodeTransientInfo {
+	return nodeTransientInfo{AllocatableVolumesCount: 0, RequestedVolumes: 0}
+}
+
+// nextGeneration: Let's make sure history never forgets the name...
+// Increments the generation number monotonically ensuring that generation numbers never collide.
+// Collision of the generation numbers would be particularly problematic if a node was deleted and
+// added back with the same name. See issue#63262.
+func nextGeneration() int64 {
+	return atomic.AddInt64(&generation, 1)
+}
+
+// nodeTransientInfo contains transient node information while scheduling.
+type nodeTransientInfo struct {
+	// AllocatableVolumesCount contains number of volumes that could be attached to node.
+	AllocatableVolumesCount int
+	// Requested number of volumes on a particular node.
+	RequestedVolumes int
+}
+
+// TransientSchedulerInfo is a transient structure which is destructed at the end of each scheduling cycle.
+// It consists of items that are valid for a scheduling cycle and is used for message passing across predicates and
+// priorities. Some examples which could be used as fields are number of volumes being used on node, current utilization
+// on node etc.
+// IMPORTANT NOTE: Make sure that each field in this structure is documented along with usage. Expand this structure
+// only when absolutely needed as this data structure will be created and destroyed during every scheduling cycle.
+type TransientSchedulerInfo struct {
+	TransientLock sync.Mutex
+	// NodeTransInfo holds the information related to nodeTransientInformation. NodeName is the key here.
+	TransNodeInfo nodeTransientInfo
+}
+
+// NewTransientSchedulerInfo returns a new scheduler transient structure with initialized values.
+func NewTransientSchedulerInfo() *TransientSchedulerInfo {
+	tsi := &TransientSchedulerInfo{
+		TransNodeInfo: initializeNodeTransientInfo(),
+	}
+	return tsi
+}
+
+// ResetTransientSchedulerInfo resets the TransientSchedulerInfo.
+func (transientSchedInfo *TransientSchedulerInfo) ResetTransientSchedulerInfo() {
+	transientSchedInfo.TransientLock.Lock()
+	defer transientSchedInfo.TransientLock.Unlock()
+	// Reset TransientNodeInfo.
+	transientSchedInfo.TransNodeInfo.AllocatableVolumesCount = 0
+	transientSchedInfo.TransNodeInfo.RequestedVolumes = 0
+}
+
+// Resource is a collection of compute resource.
+type Resource struct {
+	MilliCPU         int64
+	Memory           int64
+	EphemeralStorage int64
+	// We store allowedPodNumber (which is Node.Status.Allocatable.Pods().Value())
+	// explicitly as int, to avoid conversions and improve performance.
+	AllowedPodNumber int
+	// ScalarResources
+	ScalarResources map[v1.ResourceName]int64
+}
+
+// NewResource creates a Resource from ResourceList
+func NewResource(rl v1.ResourceList) *Resource {
+	r := &Resource{}
+	r.Add(rl)
+	return r
+}
+
+// Add adds ResourceList into Resource.
+func (r *Resource) Add(rl v1.ResourceList) {
+	if r == nil {
+		return
+	}
+
+	for rName, rQuant := range rl {
+		switch rName {
+		case v1.ResourceCPU:
+			r.MilliCPU += rQuant.MilliValue()
+		case v1.ResourceMemory:
+			r.Memory += rQuant.Value()
+		case v1.ResourcePods:
+			r.AllowedPodNumber += int(rQuant.Value())
+		case v1.ResourceEphemeralStorage:
+			r.EphemeralStorage += rQuant.Value()
+		default:
+			if v1helper.IsScalarResourceName(rName) {
+				r.AddScalar(rName, rQuant.Value())
+			}
+		}
+	}
+}
+
+// ResourceList returns a resource list of this resource.
+func (r *Resource) ResourceList() v1.ResourceList {
+	result := v1.ResourceList{
+		v1.ResourceCPU:              *resource.NewMilliQuantity(r.MilliCPU, resource.DecimalSI),
+		v1.ResourceMemory:           *resource.NewQuantity(r.Memory, resource.BinarySI),
+		v1.ResourcePods:             *resource.NewQuantity(int64(r.AllowedPodNumber), resource.BinarySI),
+		v1.ResourceEphemeralStorage: *resource.NewQuantity(r.EphemeralStorage, resource.BinarySI),
+	}
+	for rName, rQuant := range r.ScalarResources {
+		if v1helper.IsHugePageResourceName(rName) {
+			result[rName] = *resource.NewQuantity(rQuant, resource.BinarySI)
+		} else {
+			result[rName] = *resource.NewQuantity(rQuant, resource.DecimalSI)
+		}
+	}
+	return result
+}
+
+// Clone returns a copy of this resource.
+func (r *Resource) Clone() *Resource {
+	res := &Resource{
+		MilliCPU:         r.MilliCPU,
+		Memory:           r.Memory,
+		AllowedPodNumber: r.AllowedPodNumber,
+		EphemeralStorage: r.EphemeralStorage,
+	}
+	if r.ScalarResources != nil {
+		res.ScalarResources = make(map[v1.ResourceName]int64)
+		for k, v := range r.ScalarResources {
+			res.ScalarResources[k] = v
+		}
+	}
+	return res
+}
+
+// AddScalar adds a resource by a scalar value of this resource.
+func (r *Resource) AddScalar(name v1.ResourceName, quantity int64) {
+	r.SetScalar(name, r.ScalarResources[name]+quantity)
+}
+
+// SetScalar sets a resource by a scalar value of this resource.
+func (r *Resource) SetScalar(name v1.ResourceName, quantity int64) {
+	// Lazily allocate scalar resource map.
+	if r.ScalarResources == nil {
+		r.ScalarResources = map[v1.ResourceName]int64{}
+	}
+	r.ScalarResources[name] = quantity
+}
+
+// SetMaxResource compares with ResourceList and takes max value for each Resource.
+func (r *Resource) SetMaxResource(rl v1.ResourceList) {
+	if r == nil {
+		return
+	}
+
+	for rName, rQuantity := range rl {
+		switch rName {
+		case v1.ResourceMemory:
+			if mem := rQuantity.Value(); mem > r.Memory {
+				r.Memory = mem
+			}
+		case v1.ResourceCPU:
+			if cpu := rQuantity.MilliValue(); cpu > r.MilliCPU {
+				r.MilliCPU = cpu
+			}
+		case v1.ResourceEphemeralStorage:
+			if ephemeralStorage := rQuantity.Value(); ephemeralStorage > r.EphemeralStorage {
+				r.EphemeralStorage = ephemeralStorage
+			}
+		default:
+			if v1helper.IsScalarResourceName(rName) {
+				value := rQuantity.Value()
+				if value > r.ScalarResources[rName] {
+					r.SetScalar(rName, value)
+				}
+			}
+		}
+	}
+}
+
+// NewNodeInfo returns a ready to use empty NodeInfo object.
+// If any pods are given in arguments, their information will be aggregated in
+// the returned object.
+func NewNodeInfo(pods ...*v1.Pod) *NodeInfo {
+	ni := &NodeInfo{
+		requestedResource:   &Resource{},
+		nonzeroRequest:      &Resource{},
+		allocatableResource: &Resource{},
+		TransientInfo:       NewTransientSchedulerInfo(),
+		generation:          nextGeneration(),
+		usedPorts:           make(HostPortInfo),
+		imageStates:         make(map[string]*ImageStateSummary),
+	}
+	for _, pod := range pods {
+		ni.AddPod(pod)
+	}
+	return ni
+}
+
+// Node returns overall information about this node.
+func (n *NodeInfo) Node() *v1.Node {
+	if n == nil {
+		return nil
+	}
+	return n.node
+}
+
+// Pods return all pods scheduled (including assumed to be) on this node.
+func (n *NodeInfo) Pods() []*v1.Pod {
+	if n == nil {
+		return nil
+	}
+	return n.pods
+}
+
+// SetPods sets all pods scheduled (including assumed to be) on this node.
+func (n *NodeInfo) SetPods(pods []*v1.Pod) {
+	n.pods = pods
+}
+
+// UsedPorts returns used ports on this node.
+func (n *NodeInfo) UsedPorts() HostPortInfo {
+	if n == nil {
+		return nil
+	}
+	return n.usedPorts
+}
+
+// SetUsedPorts sets the used ports on this node.
+func (n *NodeInfo) SetUsedPorts(newUsedPorts HostPortInfo) {
+	n.usedPorts = newUsedPorts
+}
+
+// ImageStates returns the state information of all images.
+func (n *NodeInfo) ImageStates() map[string]*ImageStateSummary {
+	if n == nil {
+		return nil
+	}
+	return n.imageStates
+}
+
+// SetImageStates sets the state information of all images.
+func (n *NodeInfo) SetImageStates(newImageStates map[string]*ImageStateSummary) {
+	n.imageStates = newImageStates
+}
+
+// PodsWithAffinity return all pods with (anti)affinity constraints on this node.
+func (n *NodeInfo) PodsWithAffinity() []*v1.Pod {
+	if n == nil {
+		return nil
+	}
+	return n.podsWithAffinity
+}
+
+// AllowedPodNumber returns the number of the allowed pods on this node.
+func (n *NodeInfo) AllowedPodNumber() int {
+	if n == nil || n.allocatableResource == nil {
+		return 0
+	}
+	return n.allocatableResource.AllowedPodNumber
+}
+
+// Taints returns the taints list on this node.
+func (n *NodeInfo) Taints() ([]v1.Taint, error) {
+	if n == nil {
+		return nil, nil
+	}
+	return n.taints, n.taintsErr
+}
+
+// SetTaints sets the taints list on this node.
+func (n *NodeInfo) SetTaints(newTaints []v1.Taint) {
+	n.taints = newTaints
+}
+
+// MemoryPressureCondition returns the memory pressure condition status on this node.
+func (n *NodeInfo) MemoryPressureCondition() v1.ConditionStatus {
+	if n == nil {
+		return v1.ConditionUnknown
+	}
+	return n.memoryPressureCondition
+}
+
+// DiskPressureCondition returns the disk pressure condition status on this node.
+func (n *NodeInfo) DiskPressureCondition() v1.ConditionStatus {
+	if n == nil {
+		return v1.ConditionUnknown
+	}
+	return n.diskPressureCondition
+}
+
+// PIDPressureCondition returns the pid pressure condition status on this node.
+func (n *NodeInfo) PIDPressureCondition() v1.ConditionStatus {
+	if n == nil {
+		return v1.ConditionUnknown
+	}
+	return n.pidPressureCondition
+}
+
+// RequestedResource returns aggregated resource request of pods on this node.
+func (n *NodeInfo) RequestedResource() Resource {
+	if n == nil {
+		return emptyResource
+	}
+	return *n.requestedResource
+}
+
+// SetRequestedResource sets the aggregated resource request of pods on this node.
+func (n *NodeInfo) SetRequestedResource(newResource *Resource) {
+	n.requestedResource = newResource
+}
+
+// NonZeroRequest returns aggregated nonzero resource request of pods on this node.
+func (n *NodeInfo) NonZeroRequest() Resource {
+	if n == nil {
+		return emptyResource
+	}
+	return *n.nonzeroRequest
+}
+
+// SetNonZeroRequest sets the aggregated nonzero resource request of pods on this node.
+func (n *NodeInfo) SetNonZeroRequest(newResource *Resource) {
+	n.nonzeroRequest = newResource
+}
+
+// AllocatableResource returns allocatable resources on a given node.
+func (n *NodeInfo) AllocatableResource() Resource {
+	if n == nil {
+		return emptyResource
+	}
+	return *n.allocatableResource
+}
+
+// SetAllocatableResource sets the allocatableResource information of given node.
+func (n *NodeInfo) SetAllocatableResource(allocatableResource *Resource) {
+	n.allocatableResource = allocatableResource
+	n.generation = nextGeneration()
+}
+
+// GetGeneration returns the generation on this node.
+func (n *NodeInfo) GetGeneration() int64 {
+	if n == nil {
+		return 0
+	}
+	return n.generation
+}
+
+// SetGeneration sets the generation on this node. This is for testing only.
+func (n *NodeInfo) SetGeneration(newGeneration int64) {
+	n.generation = newGeneration
+}
+
+// Clone returns a copy of this node.
+func (n *NodeInfo) Clone() *NodeInfo {
+	clone := &NodeInfo{
+		node:                    n.node,
+		requestedResource:       n.requestedResource.Clone(),
+		nonzeroRequest:          n.nonzeroRequest.Clone(),
+		allocatableResource:     n.allocatableResource.Clone(),
+		taintsErr:               n.taintsErr,
+		TransientInfo:           n.TransientInfo,
+		memoryPressureCondition: n.memoryPressureCondition,
+		diskPressureCondition:   n.diskPressureCondition,
+		pidPressureCondition:    n.pidPressureCondition,
+		usedPorts:               make(HostPortInfo),
+		imageStates:             n.imageStates,
+		generation:              n.generation,
+	}
+	if len(n.pods) > 0 {
+		clone.pods = append([]*v1.Pod(nil), n.pods...)
+	}
+	if len(n.usedPorts) > 0 {
+		// HostPortInfo is a map-in-map struct
+		// make sure it's deep copied
+		for ip, portMap := range n.usedPorts {
+			clone.usedPorts[ip] = make(map[ProtocolPort]struct{})
+			for protocolPort, v := range portMap {
+				clone.usedPorts[ip][protocolPort] = v
+			}
+		}
+	}
+	if len(n.podsWithAffinity) > 0 {
+		clone.podsWithAffinity = append([]*v1.Pod(nil), n.podsWithAffinity...)
+	}
+	if len(n.taints) > 0 {
+		clone.taints = append([]v1.Taint(nil), n.taints...)
+	}
+	return clone
+}
+
+// VolumeLimits returns volume limits associated with the node
+func (n *NodeInfo) VolumeLimits() map[v1.ResourceName]int64 {
+	volumeLimits := map[v1.ResourceName]int64{}
+	for k, v := range n.AllocatableResource().ScalarResources {
+		if v1helper.IsAttachableVolumeResourceName(k) {
+			volumeLimits[k] = v
+		}
+	}
+	return volumeLimits
+}
+
+// String returns representation of human readable format of this NodeInfo.
+func (n *NodeInfo) String() string {
+	podKeys := make([]string, len(n.pods))
+	for i, pod := range n.pods {
+		podKeys[i] = pod.Name
+	}
+	return fmt.Sprintf("&NodeInfo{Pods:%v, RequestedResource:%#v, NonZeroRequest: %#v, UsedPort: %#v, AllocatableResource:%#v}",
+		podKeys, n.requestedResource, n.nonzeroRequest, n.usedPorts, n.allocatableResource)
+}
+
+func hasPodAffinityConstraints(pod *v1.Pod) bool {
+	affinity := pod.Spec.Affinity
+	return affinity != nil && (affinity.PodAffinity != nil || affinity.PodAntiAffinity != nil)
+}
+
+// AddPod adds pod information to this NodeInfo.
+func (n *NodeInfo) AddPod(pod *v1.Pod) {
+	res, non0CPU, non0Mem := calculateResource(pod)
+	n.requestedResource.MilliCPU += res.MilliCPU
+	n.requestedResource.Memory += res.Memory
+	n.requestedResource.EphemeralStorage += res.EphemeralStorage
+	if n.requestedResource.ScalarResources == nil && len(res.ScalarResources) > 0 {
+		n.requestedResource.ScalarResources = map[v1.ResourceName]int64{}
+	}
+	for rName, rQuant := range res.ScalarResources {
+		n.requestedResource.ScalarResources[rName] += rQuant
+	}
+	n.nonzeroRequest.MilliCPU += non0CPU
+	n.nonzeroRequest.Memory += non0Mem
+	n.pods = append(n.pods, pod)
+	if hasPodAffinityConstraints(pod) {
+		n.podsWithAffinity = append(n.podsWithAffinity, pod)
+	}
+
+	// Consume ports when pods added.
+	n.UpdateUsedPorts(pod, true)
+
+	n.generation = nextGeneration()
+}
+
+// RemovePod subtracts pod information from this NodeInfo.
+func (n *NodeInfo) RemovePod(pod *v1.Pod) error {
+	k1, err := GetPodKey(pod)
+	if err != nil {
+		return err
+	}
+
+	for i := range n.podsWithAffinity {
+		k2, err := GetPodKey(n.podsWithAffinity[i])
+		if err != nil {
+			klog.Errorf("Cannot get pod key, err: %v", err)
+			continue
+		}
+		if k1 == k2 {
+			// delete the element
+			n.podsWithAffinity[i] = n.podsWithAffinity[len(n.podsWithAffinity)-1]
+			n.podsWithAffinity = n.podsWithAffinity[:len(n.podsWithAffinity)-1]
+			break
+		}
+	}
+	for i := range n.pods {
+		k2, err := GetPodKey(n.pods[i])
+		if err != nil {
+			klog.Errorf("Cannot get pod key, err: %v", err)
+			continue
+		}
+		if k1 == k2 {
+			// delete the element
+			n.pods[i] = n.pods[len(n.pods)-1]
+			n.pods = n.pods[:len(n.pods)-1]
+			// reduce the resource data
+			res, non0CPU, non0Mem := calculateResource(pod)
+
+			n.requestedResource.MilliCPU -= res.MilliCPU
+			n.requestedResource.Memory -= res.Memory
+			n.requestedResource.EphemeralStorage -= res.EphemeralStorage
+			if len(res.ScalarResources) > 0 && n.requestedResource.ScalarResources == nil {
+				n.requestedResource.ScalarResources = map[v1.ResourceName]int64{}
+			}
+			for rName, rQuant := range res.ScalarResources {
+				n.requestedResource.ScalarResources[rName] -= rQuant
+			}
+			n.nonzeroRequest.MilliCPU -= non0CPU
+			n.nonzeroRequest.Memory -= non0Mem
+
+			// Release ports when remove Pods.
+			n.UpdateUsedPorts(pod, false)
+
+			n.generation = nextGeneration()
+
+			return nil
+		}
+	}
+	return fmt.Errorf("no corresponding pod %s in pods of node %s", pod.Name, n.node.Name)
+}
+
+func calculateResource(pod *v1.Pod) (res Resource, non0CPU int64, non0Mem int64) {
+	resPtr := &res
+	for _, c := range pod.Spec.Containers {
+		resPtr.Add(c.Resources.Requests)
+
+		non0CPUReq, non0MemReq := priorityutil.GetNonzeroRequests(&c.Resources.Requests)
+		non0CPU += non0CPUReq
+		non0Mem += non0MemReq
+		// No non-zero resources for GPUs or opaque resources.
+	}
+
+	return
+}
+
+// UpdateUsedPorts updates the UsedPorts of NodeInfo.
+func (n *NodeInfo) UpdateUsedPorts(pod *v1.Pod, add bool) {
+	for j := range pod.Spec.Containers {
+		container := &pod.Spec.Containers[j]
+		for k := range container.Ports {
+			podPort := &container.Ports[k]
+			if add {
+				n.usedPorts.Add(podPort.HostIP, string(podPort.Protocol), podPort.HostPort)
+			} else {
+				n.usedPorts.Remove(podPort.HostIP, string(podPort.Protocol), podPort.HostPort)
+			}
+		}
+	}
+}
+
+// SetNode sets the overall node information.
+func (n *NodeInfo) SetNode(node *v1.Node) error {
+	n.node = node
+
+	n.allocatableResource = NewResource(node.Status.Allocatable)
+
+	n.taints = node.Spec.Taints
+	for i := range node.Status.Conditions {
+		cond := &node.Status.Conditions[i]
+		switch cond.Type {
+		case v1.NodeMemoryPressure:
+			n.memoryPressureCondition = cond.Status
+		case v1.NodeDiskPressure:
+			n.diskPressureCondition = cond.Status
+		case v1.NodePIDPressure:
+			n.pidPressureCondition = cond.Status
+		default:
+			// We ignore other conditions.
+		}
+	}
+	n.TransientInfo = NewTransientSchedulerInfo()
+	n.generation = nextGeneration()
+	return nil
+}
+
+// RemoveNode removes the overall information about the node.
+func (n *NodeInfo) RemoveNode(node *v1.Node) error {
+	// We don't remove NodeInfo for because there can still be some pods on this node -
+	// this is because notifications about pods are delivered in a different watch,
+	// and thus can potentially be observed later, even though they happened before
+	// node removal. This is handled correctly in cache.go file.
+	n.node = nil
+	n.allocatableResource = &Resource{}
+	n.taints, n.taintsErr = nil, nil
+	n.memoryPressureCondition = v1.ConditionUnknown
+	n.diskPressureCondition = v1.ConditionUnknown
+	n.pidPressureCondition = v1.ConditionUnknown
+	n.imageStates = make(map[string]*ImageStateSummary)
+	n.generation = nextGeneration()
+	return nil
+}
+
+// FilterOutPods receives a list of pods and filters out those whose node names
+// are equal to the node of this NodeInfo, but are not found in the pods of this NodeInfo.
+//
+// Preemption logic simulates removal of pods on a node by removing them from the
+// corresponding NodeInfo. In order for the simulation to work, we call this method
+// on the pods returned from SchedulerCache, so that predicate functions see
+// only the pods that are not removed from the NodeInfo.
+func (n *NodeInfo) FilterOutPods(pods []*v1.Pod) []*v1.Pod {
+	node := n.Node()
+	if node == nil {
+		return pods
+	}
+	filtered := make([]*v1.Pod, 0, len(pods))
+	for _, p := range pods {
+		if p.Spec.NodeName != node.Name {
+			filtered = append(filtered, p)
+			continue
+		}
+		// If pod is on the given node, add it to 'filtered' only if it is present in nodeInfo.
+		podKey, _ := GetPodKey(p)
+		for _, np := range n.Pods() {
+			npodkey, _ := GetPodKey(np)
+			if npodkey == podKey {
+				filtered = append(filtered, p)
+				break
+			}
+		}
+	}
+	return filtered
+}
+
+// GetPodKey returns the string key of a pod.
+func GetPodKey(pod *v1.Pod) (string, error) {
+	uid := string(pod.UID)
+	if len(uid) == 0 {
+		return "", errors.New("Cannot get cache key for pod with empty UID")
+	}
+	return uid, nil
+}
+
+// Filter implements PodFilter interface. It returns false only if the pod node name
+// matches NodeInfo.node and the pod is not found in the pods list. Otherwise,
+// returns true.
+func (n *NodeInfo) Filter(pod *v1.Pod) bool {
+	if pod.Spec.NodeName != n.node.Name {
+		return true
+	}
+	for _, p := range n.pods {
+		if p.Name == pod.Name && p.Namespace == pod.Namespace {
+			return true
+		}
+	}
+	return false
+}

vendor/k8s.io/kubernetes/pkg/scheduler/nodeinfo/util.go

@@ -0,0 +1,78 @@
+/*
+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 nodeinfo
+
+import (
+	"k8s.io/api/core/v1"
+	"k8s.io/apimachinery/pkg/util/sets"
+)
+
+// CreateNodeNameToInfoMap obtains a list of pods and pivots that list into a map where the keys are node names
+// and the values are the aggregated information for that node.
+func CreateNodeNameToInfoMap(pods []*v1.Pod, nodes []*v1.Node) map[string]*NodeInfo {
+	nodeNameToInfo := make(map[string]*NodeInfo)
+	for _, pod := range pods {
+		nodeName := pod.Spec.NodeName
+		if _, ok := nodeNameToInfo[nodeName]; !ok {
+			nodeNameToInfo[nodeName] = NewNodeInfo()
+		}
+		nodeNameToInfo[nodeName].AddPod(pod)
+	}
+	imageExistenceMap := createImageExistenceMap(nodes)
+
+	for _, node := range nodes {
+		if _, ok := nodeNameToInfo[node.Name]; !ok {
+			nodeNameToInfo[node.Name] = NewNodeInfo()
+		}
+		nodeInfo := nodeNameToInfo[node.Name]
+		nodeInfo.SetNode(node)
+		nodeInfo.imageStates = getNodeImageStates(node, imageExistenceMap)
+	}
+	return nodeNameToInfo
+}
+
+// getNodeImageStates returns the given node's image states based on the given imageExistence map.
+func getNodeImageStates(node *v1.Node, imageExistenceMap map[string]sets.String) map[string]*ImageStateSummary {
+	imageStates := make(map[string]*ImageStateSummary)
+
+	for _, image := range node.Status.Images {
+		for _, name := range image.Names {
+			imageStates[name] = &ImageStateSummary{
+				Size:     image.SizeBytes,
+				NumNodes: len(imageExistenceMap[name]),
+			}
+		}
+	}
+	return imageStates
+}
+
+// createImageExistenceMap returns a map recording on which nodes the images exist, keyed by the images' names.
+func createImageExistenceMap(nodes []*v1.Node) map[string]sets.String {
+	imageExistenceMap := make(map[string]sets.String)
+	for _, node := range nodes {
+		for _, image := range node.Status.Images {
+			for _, name := range image.Names {
+				if _, ok := imageExistenceMap[name]; !ok {
+					imageExistenceMap[name] = sets.NewString(node.Name)
+				} else {
+					imageExistenceMap[name].Insert(node.Name)
+				}
+			}
+		}
+	}
+	return imageExistenceMap
+}

vendor/k8s.io/kubernetes/pkg/scheduler/util/backoff_utils.go

@@ -0,0 +1,220 @@
+/*
+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 util
+
+import (
+	"fmt"
+	"sync"
+	"sync/atomic"
+	"time"
+
+	ktypes "k8s.io/apimachinery/pkg/types"
+
+	"k8s.io/klog"
+)
+
+type clock interface {
+	Now() time.Time
+}
+
+type realClock struct{}
+
+func (realClock) Now() time.Time {
+	return time.Now()
+}
+
+// backoffEntry is single threaded.  in particular, it only allows a single action to be waiting on backoff at a time.
+// It is also not safe to copy this object.
+type backoffEntry struct {
+	initialized bool
+	podName     ktypes.NamespacedName
+	backoff     time.Duration
+	lastUpdate  time.Time
+	reqInFlight int32
+}
+
+// tryLock attempts to acquire a lock via atomic compare and swap.
+// returns true if the lock was acquired, false otherwise
+func (b *backoffEntry) tryLock() bool {
+	return atomic.CompareAndSwapInt32(&b.reqInFlight, 0, 1)
+}
+
+// unlock returns the lock.  panics if the lock isn't held
+func (b *backoffEntry) unlock() {
+	if !atomic.CompareAndSwapInt32(&b.reqInFlight, 1, 0) {
+		panic(fmt.Sprintf("unexpected state on unlocking: %+v", b))
+	}
+}
+
+// backoffTime returns the Time when a backoffEntry completes backoff
+func (b *backoffEntry) backoffTime() time.Time {
+	return b.lastUpdate.Add(b.backoff)
+}
+
+// getBackoff returns the duration until this entry completes backoff
+func (b *backoffEntry) getBackoff(maxDuration time.Duration) time.Duration {
+	if !b.initialized {
+		b.initialized = true
+		return b.backoff
+	}
+	newDuration := b.backoff * 2
+	if newDuration > maxDuration {
+		newDuration = maxDuration
+	}
+	b.backoff = newDuration
+	klog.V(4).Infof("Backing off %s", newDuration.String())
+	return newDuration
+}
+
+// PodBackoff is used to restart a pod with back-off delay.
+type PodBackoff struct {
+	// expiryQ stores backoffEntry orderedy by lastUpdate until they reach maxDuration and are GC'd
+	expiryQ         *Heap
+	lock            sync.Mutex
+	clock           clock
+	defaultDuration time.Duration
+	maxDuration     time.Duration
+}
+
+// MaxDuration returns the max time duration of the back-off.
+func (p *PodBackoff) MaxDuration() time.Duration {
+	return p.maxDuration
+}
+
+// CreateDefaultPodBackoff creates a default pod back-off object.
+func CreateDefaultPodBackoff() *PodBackoff {
+	return CreatePodBackoff(1*time.Second, 60*time.Second)
+}
+
+// CreatePodBackoff creates a pod back-off object by default duration and max duration.
+func CreatePodBackoff(defaultDuration, maxDuration time.Duration) *PodBackoff {
+	return CreatePodBackoffWithClock(defaultDuration, maxDuration, realClock{})
+}
+
+// CreatePodBackoffWithClock creates a pod back-off object by default duration, max duration and clock.
+func CreatePodBackoffWithClock(defaultDuration, maxDuration time.Duration, clock clock) *PodBackoff {
+	return &PodBackoff{
+		expiryQ:         NewHeap(backoffEntryKeyFunc, backoffEntryCompareUpdate),
+		clock:           clock,
+		defaultDuration: defaultDuration,
+		maxDuration:     maxDuration,
+	}
+}
+
+// getEntry returns the backoffEntry for a given podID
+func (p *PodBackoff) getEntry(podID ktypes.NamespacedName) *backoffEntry {
+	entry, exists, _ := p.expiryQ.GetByKey(podID.String())
+	var be *backoffEntry
+	if !exists {
+		be = &backoffEntry{
+			initialized: false,
+			podName:     podID,
+			backoff:     p.defaultDuration,
+		}
+		p.expiryQ.Update(be)
+	} else {
+		be = entry.(*backoffEntry)
+	}
+	return be
+}
+
+// BackoffPod updates the backoff for a podId and returns the duration until backoff completion
+func (p *PodBackoff) BackoffPod(podID ktypes.NamespacedName) time.Duration {
+	p.lock.Lock()
+	defer p.lock.Unlock()
+	entry := p.getEntry(podID)
+	entry.lastUpdate = p.clock.Now()
+	p.expiryQ.Update(entry)
+	return entry.getBackoff(p.maxDuration)
+}
+
+// TryBackoffAndWait tries to acquire the backoff lock
+func (p *PodBackoff) TryBackoffAndWait(podID ktypes.NamespacedName, stop <-chan struct{}) bool {
+	p.lock.Lock()
+	entry := p.getEntry(podID)
+
+	if !entry.tryLock() {
+		p.lock.Unlock()
+		return false
+	}
+	defer entry.unlock()
+	duration := entry.getBackoff(p.maxDuration)
+	p.lock.Unlock()
+	select {
+	case <-time.After(duration):
+		return true
+	case <-stop:
+		return false
+	}
+}
+
+// Gc execute garbage collection on the pod back-off.
+func (p *PodBackoff) Gc() {
+	p.lock.Lock()
+	defer p.lock.Unlock()
+	now := p.clock.Now()
+	var be *backoffEntry
+	for {
+		entry := p.expiryQ.Peek()
+		if entry == nil {
+			break
+		}
+		be = entry.(*backoffEntry)
+		if now.Sub(be.lastUpdate) > p.maxDuration {
+			p.expiryQ.Pop()
+		} else {
+			break
+		}
+	}
+}
+
+// GetBackoffTime returns the time that podID completes backoff
+func (p *PodBackoff) GetBackoffTime(podID ktypes.NamespacedName) (time.Time, bool) {
+	p.lock.Lock()
+	defer p.lock.Unlock()
+	rawBe, exists, _ := p.expiryQ.GetByKey(podID.String())
+	if !exists {
+		return time.Time{}, false
+	}
+	be := rawBe.(*backoffEntry)
+	return be.lastUpdate.Add(be.backoff), true
+}
+
+// ClearPodBackoff removes all tracking information for podID (clears expiry)
+func (p *PodBackoff) ClearPodBackoff(podID ktypes.NamespacedName) bool {
+	p.lock.Lock()
+	defer p.lock.Unlock()
+	entry, exists, _ := p.expiryQ.GetByKey(podID.String())
+	if exists {
+		err := p.expiryQ.Delete(entry)
+		return err == nil
+	}
+	return false
+}
+
+// backoffEntryKeyFunc is the keying function used for mapping a backoffEntry to string for heap
+func backoffEntryKeyFunc(b interface{}) (string, error) {
+	be := b.(*backoffEntry)
+	return be.podName.String(), nil
+}
+
+// backoffEntryCompareUpdate returns true when b1's backoff time is before b2's
+func backoffEntryCompareUpdate(b1, b2 interface{}) bool {
+	be1 := b1.(*backoffEntry)
+	be2 := b2.(*backoffEntry)
+	return be1.lastUpdate.Before(be2.lastUpdate)
+}

vendor/k8s.io/kubernetes/pkg/scheduler/util/clock.go

@@ -0,0 +1,34 @@
+/*
+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 util
+
+import (
+	"time"
+)
+
+// Clock provides an interface for getting the current time
+type Clock interface {
+	Now() time.Time
+}
+
+// RealClock implements a clock using time
+type RealClock struct{}
+
+// Now returns the current time with time.Now
+func (RealClock) Now() time.Time {
+	return time.Now()
+}

vendor/k8s.io/kubernetes/pkg/scheduler/util/heap.go

@@ -0,0 +1,236 @@
+/*
+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.
+*/
+
+// Below is the implementation of the a heap. The logic is pretty much the same
+// as cache.heap, however, this heap does not perform synchronization. It leaves
+// synchronization to the SchedulingQueue.
+
+package util
+
+import (
+	"container/heap"
+	"fmt"
+
+	"k8s.io/client-go/tools/cache"
+)
+
+// KeyFunc is a function type to get the key from an object.
+type KeyFunc func(obj interface{}) (string, error)
+
+type heapItem struct {
+	obj   interface{} // The object which is stored in the heap.
+	index int         // The index of the object's key in the Heap.queue.
+}
+
+type itemKeyValue struct {
+	key string
+	obj interface{}
+}
+
+// heapData is an internal struct that implements the standard heap interface
+// and keeps the data stored in the heap.
+type heapData struct {
+	// items is a map from key of the objects to the objects and their index.
+	// We depend on the property that items in the map are in the queue and vice versa.
+	items map[string]*heapItem
+	// queue implements a heap data structure and keeps the order of elements
+	// according to the heap invariant. The queue keeps the keys of objects stored
+	// in "items".
+	queue []string
+
+	// keyFunc is used to make the key used for queued item insertion and retrieval, and
+	// should be deterministic.
+	keyFunc KeyFunc
+	// lessFunc is used to compare two objects in the heap.
+	lessFunc LessFunc
+}
+
+var (
+	_ = heap.Interface(&heapData{}) // heapData is a standard heap
+)
+
+// Less compares two objects and returns true if the first one should go
+// in front of the second one in the heap.
+func (h *heapData) Less(i, j int) bool {
+	if i > len(h.queue) || j > len(h.queue) {
+		return false
+	}
+	itemi, ok := h.items[h.queue[i]]
+	if !ok {
+		return false
+	}
+	itemj, ok := h.items[h.queue[j]]
+	if !ok {
+		return false
+	}
+	return h.lessFunc(itemi.obj, itemj.obj)
+}
+
+// Len returns the number of items in the Heap.
+func (h *heapData) Len() int { return len(h.queue) }
+
+// Swap implements swapping of two elements in the heap. This is a part of standard
+// heap interface and should never be called directly.
+func (h *heapData) Swap(i, j int) {
+	h.queue[i], h.queue[j] = h.queue[j], h.queue[i]
+	item := h.items[h.queue[i]]
+	item.index = i
+	item = h.items[h.queue[j]]
+	item.index = j
+}
+
+// Push is supposed to be called by heap.Push only.
+func (h *heapData) Push(kv interface{}) {
+	keyValue := kv.(*itemKeyValue)
+	n := len(h.queue)
+	h.items[keyValue.key] = &heapItem{keyValue.obj, n}
+	h.queue = append(h.queue, keyValue.key)
+}
+
+// Pop is supposed to be called by heap.Pop only.
+func (h *heapData) Pop() interface{} {
+	key := h.queue[len(h.queue)-1]
+	h.queue = h.queue[0 : len(h.queue)-1]
+	item, ok := h.items[key]
+	if !ok {
+		// This is an error
+		return nil
+	}
+	delete(h.items, key)
+	return item.obj
+}
+
+// Peek is supposed to be called by heap.Peek only.
+func (h *heapData) Peek() interface{} {
+	if len(h.queue) > 0 {
+		return h.items[h.queue[0]].obj
+	}
+	return nil
+}
+
+// Heap is a producer/consumer queue that implements a heap data structure.
+// It can be used to implement priority queues and similar data structures.
+type Heap struct {
+	// data stores objects and has a queue that keeps their ordering according
+	// to the heap invariant.
+	data *heapData
+}
+
+// Add inserts an item, and puts it in the queue. The item is updated if it
+// already exists.
+func (h *Heap) Add(obj interface{}) error {
+	key, err := h.data.keyFunc(obj)
+	if err != nil {
+		return cache.KeyError{Obj: obj, Err: err}
+	}
+	if _, exists := h.data.items[key]; exists {
+		h.data.items[key].obj = obj
+		heap.Fix(h.data, h.data.items[key].index)
+	} else {
+		heap.Push(h.data, &itemKeyValue{key, obj})
+	}
+	return nil
+}
+
+// AddIfNotPresent inserts an item, and puts it in the queue. If an item with
+// the key is present in the map, no changes is made to the item.
+func (h *Heap) AddIfNotPresent(obj interface{}) error {
+	key, err := h.data.keyFunc(obj)
+	if err != nil {
+		return cache.KeyError{Obj: obj, Err: err}
+	}
+	if _, exists := h.data.items[key]; !exists {
+		heap.Push(h.data, &itemKeyValue{key, obj})
+	}
+	return nil
+}
+
+// Update is the same as Add in this implementation. When the item does not
+// exist, it is added.
+func (h *Heap) Update(obj interface{}) error {
+	return h.Add(obj)
+}
+
+// Delete removes an item.
+func (h *Heap) Delete(obj interface{}) error {
+	key, err := h.data.keyFunc(obj)
+	if err != nil {
+		return cache.KeyError{Obj: obj, Err: err}
+	}
+	if item, ok := h.data.items[key]; ok {
+		heap.Remove(h.data, item.index)
+		return nil
+	}
+	return fmt.Errorf("object not found")
+}
+
+// Peek returns the head of the heap without removing it.
+func (h *Heap) Peek() interface{} {
+	return h.data.Peek()
+}
+
+// Pop returns the head of the heap and removes it.
+func (h *Heap) Pop() (interface{}, error) {
+	obj := heap.Pop(h.data)
+	if obj != nil {
+		return obj, nil
+	}
+	return nil, fmt.Errorf("object was removed from heap data")
+}
+
+// Get returns the requested item, or sets exists=false.
+func (h *Heap) Get(obj interface{}) (interface{}, bool, error) {
+	key, err := h.data.keyFunc(obj)
+	if err != nil {
+		return nil, false, cache.KeyError{Obj: obj, Err: err}
+	}
+	return h.GetByKey(key)
+}
+
+// GetByKey returns the requested item, or sets exists=false.
+func (h *Heap) GetByKey(key string) (interface{}, bool, error) {
+	item, exists := h.data.items[key]
+	if !exists {
+		return nil, false, nil
+	}
+	return item.obj, true, nil
+}
+
+// List returns a list of all the items.
+func (h *Heap) List() []interface{} {
+	list := make([]interface{}, 0, len(h.data.items))
+	for _, item := range h.data.items {
+		list = append(list, item.obj)
+	}
+	return list
+}
+
+// Len returns the number of items in the heap.
+func (h *Heap) Len() int {
+	return len(h.data.queue)
+}
+
+// NewHeap returns a Heap which can be used to queue up items to process.
+func NewHeap(keyFn KeyFunc, lessFn LessFunc) *Heap {
+	return &Heap{
+		data: &heapData{
+			items:    map[string]*heapItem{},
+			queue:    []string{},
+			keyFunc:  keyFn,
+			lessFunc: lessFn,
+		},
+	}
+}

vendor/k8s.io/kubernetes/pkg/scheduler/util/utils.go

@@ -0,0 +1,99 @@
+/*
+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 util
+
+import (
+	"sort"
+
+	"k8s.io/api/core/v1"
+	"k8s.io/apiserver/pkg/util/feature"
+	"k8s.io/kubernetes/pkg/apis/scheduling"
+	"k8s.io/kubernetes/pkg/features"
+)
+
+// GetContainerPorts returns the used host ports of Pods: if 'port' was used, a 'port:true' pair
+// will be in the result; but it does not resolve port conflict.
+func GetContainerPorts(pods ...*v1.Pod) []*v1.ContainerPort {
+	var ports []*v1.ContainerPort
+	for _, pod := range pods {
+		for j := range pod.Spec.Containers {
+			container := &pod.Spec.Containers[j]
+			for k := range container.Ports {
+				ports = append(ports, &container.Ports[k])
+			}
+		}
+	}
+	return ports
+}
+
+// PodPriorityEnabled indicates whether pod priority feature is enabled.
+func PodPriorityEnabled() bool {
+	return feature.DefaultFeatureGate.Enabled(features.PodPriority)
+}
+
+// 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).
+	return pod.Name + "_" + pod.Namespace
+}
+
+// GetPodPriority return 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 scheduling.DefaultPriorityWhenNoDefaultClassExists
+}
+
+// SortableList is a list that implements sort.Interface.
+type SortableList struct {
+	Items    []interface{}
+	CompFunc LessFunc
+}
+
+// LessFunc is a function that receives two items and returns true if the first
+// item should be placed before the second one when the list is sorted.
+type LessFunc func(item1, item2 interface{}) bool
+
+var _ = sort.Interface(&SortableList{})
+
+func (l *SortableList) Len() int { return len(l.Items) }
+
+func (l *SortableList) Less(i, j int) bool {
+	return l.CompFunc(l.Items[i], l.Items[j])
+}
+
+func (l *SortableList) Swap(i, j int) {
+	l.Items[i], l.Items[j] = l.Items[j], l.Items[i]
+}
+
+// Sort sorts the items in the list using the given CompFunc. Item1 is placed
+// before Item2 when CompFunc(Item1, Item2) returns true.
+func (l *SortableList) Sort() {
+	sort.Sort(l)
+}
+
+// HigherPriorityPod return true when priority of the first pod is higher than
+// the second one. It takes arguments of the type "interface{}" to be used with
+// SortableList, but expects those arguments to be *v1.Pod.
+func HigherPriorityPod(pod1, pod2 interface{}) bool {
+	return GetPodPriority(pod1.(*v1.Pod)) > GetPodPriority(pod2.(*v1.Pod))
+}

vendor/k8s.io/kubernetes/pkg/scheduler/volumebinder/volume_binder.go

@@ -0,0 +1,61 @@
+/*
+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 volumebinder
+
+import (
+	"time"
+
+	"k8s.io/api/core/v1"
+	coreinformers "k8s.io/client-go/informers/core/v1"
+	storageinformers "k8s.io/client-go/informers/storage/v1"
+	clientset "k8s.io/client-go/kubernetes"
+	"k8s.io/kubernetes/pkg/controller/volume/persistentvolume"
+)
+
+// VolumeBinder sets up the volume binding library
+type VolumeBinder struct {
+	Binder persistentvolume.SchedulerVolumeBinder
+}
+
+// NewVolumeBinder sets up the volume binding library and binding queue
+func NewVolumeBinder(
+	client clientset.Interface,
+	nodeInformer coreinformers.NodeInformer,
+	pvcInformer coreinformers.PersistentVolumeClaimInformer,
+	pvInformer coreinformers.PersistentVolumeInformer,
+	storageClassInformer storageinformers.StorageClassInformer,
+	bindTimeout time.Duration) *VolumeBinder {
+
+	return &VolumeBinder{
+		Binder: persistentvolume.NewVolumeBinder(client, nodeInformer, pvcInformer, pvInformer, storageClassInformer, bindTimeout),
+	}
+}
+
+// NewFakeVolumeBinder sets up a fake volume binder and binding queue
+func NewFakeVolumeBinder(config *persistentvolume.FakeVolumeBinderConfig) *VolumeBinder {
+	return &VolumeBinder{
+		Binder: persistentvolume.NewFakeVolumeBinder(config),
+	}
+}
+
+// DeletePodBindings will delete the cached volume bindings for the given pod.
+func (b *VolumeBinder) DeletePodBindings(pod *v1.Pod) {
+	cache := b.Binder.GetBindingsCache()
+	if cache != nil && pod != nil {
+		cache.DeleteBindings(pod)
+	}
+}