// Calculates score for all pods and returns podInfo structure.
// Score is defined as cpu_sum/node_capacity + mem_sum/node_capacity.
// Pods that have bigger requirements should be processed first, thus have higher scores.
func calculatePodScore(pods []*apiv1.Pod, nodeTemplate *schedulercache.NodeInfo) []*podInfo {
podInfos := make([]*podInfo, 0, len(pods))
for _, pod := range pods {
cpuSum := resource.Quantity{}
memorySum := resource.Quantity{}
for _, container := range pod.Spec.Containers {
if request, ok := container.Resources.Requests[apiv1.ResourceCPU]; ok {
cpuSum.Add(request)
}
if request, ok := container.Resources.Requests[apiv1.ResourceMemory]; ok {
memorySum.Add(request)
}
}
score := float64(0)
if cpuAllocatable, ok := nodeTemplate.Node().Status.Allocatable[apiv1.ResourceCPU]; ok && cpuAllocatable.MilliValue() > 0 {
score += float64(cpuSum.MilliValue()) / float64(cpuAllocatable.MilliValue())
}
if memAllocatable, ok := nodeTemplate.Node().Status.Allocatable[apiv1.ResourceMemory]; ok && memAllocatable.Value() > 0 {
score += float64(memorySum.Value()) / float64(memAllocatable.Value())
}
podInfos = append(podInfos, &podInfo{
score: score,
pod: pod,
})
}
return podInfos
}
// ComputeTaintTolerationPriority prepares the priority list for all the nodes based on the number of intolerable taints on the node
func ComputeTaintTolerationPriorityMap(pod *api.Pod, meta interface{}, nodeInfo *schedulercache.NodeInfo) (schedulerapi.HostPriority, error) {
node := nodeInfo.Node()
if node == nil {
return schedulerapi.HostPriority{}, fmt.Errorf("node not found")
}
var tolerationList []api.Toleration
if priorityMeta, ok := meta.(*priorityMetadata); ok {
tolerationList = priorityMeta.podTolerations
} else {
var err error
tolerationList, err = getTolerationListFromPod(pod)
if err != nil {
return schedulerapi.HostPriority{}, err
}
}
taints, err := api.GetTaintsFromNodeAnnotations(node.Annotations)
if err != nil {
return schedulerapi.HostPriority{}, err
}
return schedulerapi.HostPriority{
Host: node.Name,
Score: countIntolerableTaintsPreferNoSchedule(taints, tolerationList),
}, nil
}
func CalculateNodePreferAvoidPodsPriorityMap(pod *v1.Pod, meta interface{}, nodeInfo *schedulercache.NodeInfo) (schedulerapi.HostPriority, error) {
node := nodeInfo.Node()
if node == nil {
return schedulerapi.HostPriority{}, fmt.Errorf("node not found")
}
controllerRef := priorityutil.GetControllerRef(pod)
if controllerRef != nil {
// Ignore pods that are owned by other controller than ReplicationController
// or ReplicaSet.
if controllerRef.Kind != "ReplicationController" && controllerRef.Kind != "ReplicaSet" {
controllerRef = nil
}
}
if controllerRef == nil {
return schedulerapi.HostPriority{Host: node.Name, Score: 10}, nil
}
avoids, err := v1.GetAvoidPodsFromNodeAnnotations(node.Annotations)
if err != nil {
// If we cannot get annotation, assume it's schedulable there.
return schedulerapi.HostPriority{Host: node.Name, Score: 10}, nil
}
for i := range avoids.PreferAvoidPods {
avoid := &avoids.PreferAvoidPods[i]
if controllerRef != nil {
if avoid.PodSignature.PodController.Kind == controllerRef.Kind && avoid.PodSignature.PodController.UID == controllerRef.UID {
return schedulerapi.HostPriority{Host: node.Name, Score: 0}, nil
}
}
}
return schedulerapi.HostPriority{Host: node.Name, Score: 10}, nil
}
func (checker *PodAffinityChecker) InterPodAffinityMatches(pod *api.Pod, meta interface{}, nodeInfo *schedulercache.NodeInfo) (bool, error) {
node := nodeInfo.Node()
if node == nil {
return false, fmt.Errorf("node not found")
}
allPods, err := checker.podLister.List(labels.Everything())
if err != nil {
return false, err
}
affinity, err := api.GetAffinityFromPodAnnotations(pod.Annotations)
if err != nil {
return false, err
}
// Check if the current node match the inter-pod affinity scheduling constraints.
// Hard inter-pod affinity is not symmetric, check only when affinity.PodAffinity exists.
if affinity.PodAffinity != nil {
if !checker.NodeMatchesHardPodAffinity(pod, allPods, node, affinity.PodAffinity) {
return false, ErrPodAffinityNotMatch
}
}
// Hard inter-pod anti-affinity is symmetric, we should always check it.
if !checker.NodeMatchesHardPodAntiAffinity(pod, allPods, node, affinity.PodAntiAffinity) {
return false, ErrPodAffinityNotMatch
}
return true, nil
}
// Calculate the resource used on a node. 'node' has information about the resources on the node.
// 'pods' is a list of pods currently scheduled on the node.
func calculateUsedPriority(pod *api.Pod, podRequests *schedulercache.Resource, nodeInfo *schedulercache.NodeInfo) (schedulerapi.HostPriority, error) {
node := nodeInfo.Node()
if node == nil {
return schedulerapi.HostPriority{}, fmt.Errorf("node not found")
}
allocatableResources := nodeInfo.AllocatableResource()
totalResources := *podRequests
totalResources.MilliCPU += nodeInfo.NonZeroRequest().MilliCPU
totalResources.Memory += nodeInfo.NonZeroRequest().Memory
cpuScore := calculateUsedScore(totalResources.MilliCPU, allocatableResources.MilliCPU, node.Name)
memoryScore := calculateUsedScore(totalResources.Memory, allocatableResources.Memory, node.Name)
if glog.V(10) {
// We explicitly don't do glog.V(10).Infof() to avoid computing all the parameters if this is
// not logged. There is visible performance gain from it.
glog.V(10).Infof(
"%v -> %v: Most Requested Priority, capacity %d millicores %d memory bytes, total request %d millicores %d memory bytes, score %d CPU %d memory",
pod.Name, node.Name,
allocatableResources.MilliCPU, allocatableResources.Memory,
totalResources.MilliCPU, totalResources.Memory,
cpuScore, memoryScore,
)
}
return schedulerapi.HostPriority{
Host: node.Name,
Score: int((cpuScore + memoryScore) / 2),
}, nil
}
func PodFitsResources(pod *api.Pod, nodeInfo *schedulercache.NodeInfo) (bool, error) {
node := nodeInfo.Node()
if node == nil {
return false, fmt.Errorf("node not found")
}
allocatable := node.Status.Allocatable
allowedPodNumber := allocatable.Pods().Value()
if int64(len(nodeInfo.Pods()))+1 > allowedPodNumber {
return false,
newInsufficientResourceError(podCountResourceName, 1, int64(len(nodeInfo.Pods())), allowedPodNumber)
}
podRequest := getResourceRequest(pod)
if podRequest.milliCPU == 0 && podRequest.memory == 0 {
return true, nil
}
totalMilliCPU := allocatable.Cpu().MilliValue()
totalMemory := allocatable.Memory().Value()
if totalMilliCPU < podRequest.milliCPU+nodeInfo.RequestedResource().MilliCPU {
return false,
newInsufficientResourceError(cpuResourceName, podRequest.milliCPU, nodeInfo.RequestedResource().MilliCPU, totalMilliCPU)
}
if totalMemory < podRequest.memory+nodeInfo.RequestedResource().Memory {
return false,
newInsufficientResourceError(memoryResoureceName, podRequest.memory, nodeInfo.RequestedResource().Memory, totalMemory)
}
glog.V(10).Infof("Schedule Pod %+v on Node %+v is allowed, Node is running only %v out of %v Pods.",
podName(pod), node.Name, len(nodeInfo.Pods()), allowedPodNumber)
return true, nil
}
// CheckNodeMemoryPressurePredicate checks if a pod can be scheduled on a node
// reporting memory pressure condition.
func CheckNodeMemoryPressurePredicate(pod *api.Pod, meta interface{}, nodeInfo *schedulercache.NodeInfo) (bool, error) {
node := nodeInfo.Node()
if node == nil {
return false, fmt.Errorf("node not found")
}
var podBestEffort bool
predicateMeta, ok := meta.(*predicateMetadata)
if ok {
podBestEffort = predicateMeta.podBestEffort
} else {
// We couldn't parse metadata - fallback to computing it.
podBestEffort = isPodBestEffort(pod)
}
// pod is not BestEffort pod
if !podBestEffort {
return true, nil
}
// is node under presure?
for _, cond := range node.Status.Conditions {
if cond.Type == api.NodeMemoryPressure && cond.Status == api.ConditionTrue {
return false, ErrNodeUnderMemoryPressure
}
}
return true, nil
}
func (c *PodAffinityChecker) InterPodAffinityMatches(pod *api.Pod, meta interface{}, nodeInfo *schedulercache.NodeInfo) (bool, []algorithm.PredicateFailureReason, error) {
node := nodeInfo.Node()
if node == nil {
return false, nil, fmt.Errorf("node not found")
}
if !c.satisfiesExistingPodsAntiAffinity(pod, meta, node) {
return false, []algorithm.PredicateFailureReason{ErrPodAffinityNotMatch}, nil
}
// Now check if <pod> requirements will be satisfied on this node.
affinity, err := api.GetAffinityFromPodAnnotations(pod.Annotations)
if err != nil {
return false, nil, err
}
if affinity == nil || (affinity.PodAffinity == nil && affinity.PodAntiAffinity == nil) {
return true, nil, nil
}
if !c.satisfiesPodsAffinityAntiAffinity(pod, node, affinity) {
return false, []algorithm.PredicateFailureReason{ErrPodAffinityNotMatch}, nil
}
if glog.V(10) {
// We explicitly don't do glog.V(10).Infof() to avoid computing all the parameters if this is
// not logged. There is visible performance gain from it.
glog.Infof("Schedule Pod %+v on Node %+v is allowed, pod (anti)affinity constraints satisfied",
podName(pod), node.Name)
}
return true, nil, nil
}
func PodSelectorMatches(pod *api.Pod, meta interface{}, nodeInfo *schedulercache.NodeInfo) (bool, []algorithm.PredicateFailureReason, error) {
node := nodeInfo.Node()
if node == nil {
return false, nil, fmt.Errorf("node not found")
}
if podMatchesNodeLabels(pod, node) {
return true, nil, nil
}
return false, []algorithm.PredicateFailureReason{ErrNodeSelectorNotMatch}, nil
}
func matchesPredicate(pod *api.Pod, nodeInfo *schedulercache.NodeInfo) (bool, error) {
node := nodeInfo.Node()
if node == nil {
return false, fmt.Errorf("node not found")
}
if pod.Name == node.Name {
return true, nil
}
return false, algorithmpredicates.ErrFakePredicate
}
func PodSelectorMatches(pod *api.Pod, nodeName string, nodeInfo *schedulercache.NodeInfo) (bool, error) {
node := nodeInfo.Node()
if node == nil {
return false, fmt.Errorf("node not found: %q", nodeName)
}
if PodMatchesNodeLabels(pod, node) {
return true, nil
}
return false, ErrNodeSelectorNotMatch
}
func PodSelectorMatches(pod *api.Pod, meta interface{}, nodeInfo *schedulercache.NodeInfo) (bool, error) {
node := nodeInfo.Node()
if node == nil {
return false, fmt.Errorf("node not found")
}
if podMatchesNodeLabels(pod, node) {
return true, nil
}
return false, ErrNodeSelectorNotMatch
}
func matchesPredicate(pod *api.Pod, meta interface{}, nodeInfo *schedulercache.NodeInfo) (bool, []algorithm.PredicateFailureReason, error) {
node := nodeInfo.Node()
if node == nil {
return false, nil, fmt.Errorf("node not found")
}
if pod.Name == node.Name {
return true, nil, nil
}
return false, []algorithm.PredicateFailureReason{algorithmpredicates.ErrFakePredicate}, nil
}
func PodFitsHost(pod *v1.Pod, meta interface{}, nodeInfo *schedulercache.NodeInfo) (bool, []algorithm.PredicateFailureReason, error) {
if len(pod.Spec.NodeName) == 0 {
return true, nil, nil
}
node := nodeInfo.Node()
if node == nil {
return false, nil, fmt.Errorf("node not found")
}
if pod.Spec.NodeName == node.Name {
return true, nil, nil
}
return false, []algorithm.PredicateFailureReason{ErrPodNotMatchHostName}, nil
}
func PodFitsHost(pod *api.Pod, meta interface{}, nodeInfo *schedulercache.NodeInfo) (bool, error) {
if len(pod.Spec.NodeName) == 0 {
return true, nil
}
node := nodeInfo.Node()
if node == nil {
return false, fmt.Errorf("node not found")
}
if pod.Spec.NodeName == node.Name {
return true, nil
}
return false, ErrPodNotMatchHostName
}
func (checker *PodAffinityChecker) InterPodAffinityMatches(pod *api.Pod, nodeInfo *schedulercache.NodeInfo) (bool, error) {
node := nodeInfo.Node()
if node == nil {
return false, fmt.Errorf("node not found")
}
allPods, err := checker.podLister.List(labels.Everything())
if err != nil {
return false, err
}
if checker.NodeMatchPodAffinityAntiAffinity(pod, allPods, node) {
return true, nil
}
return false, ErrPodAffinityNotMatch
}
// CheckNodeInodePressurePredicate checks if a pod can be scheduled on a node
// reporting inode pressure condition.
func CheckNodeInodePressurePredicate(pod *api.Pod, meta interface{}, nodeInfo *schedulercache.NodeInfo) (bool, []algorithm.PredicateFailureReason, error) {
node := nodeInfo.Node()
if node == nil {
return false, nil, fmt.Errorf("node not found")
}
// is node under presure?
for _, cond := range node.Status.Conditions {
if cond.Type == api.NodeInodePressure && cond.Status == api.ConditionTrue {
return false, []algorithm.PredicateFailureReason{ErrNodeUnderInodePressure}, nil
}
}
return true, nil, nil
}
// ImageLocalityPriority is a priority function that favors nodes that already have requested pod container's images.
// It will detect whether the requested images are present on a node, and then calculate a score ranging from 0 to 10
// based on the total size of those images.
// - If none of the images are present, this node will be given the lowest priority.
// - If some of the images are present on a node, the larger their sizes' sum, the higher the node's priority.
func ImageLocalityPriorityMap(pod *api.Pod, meta interface{}, nodeInfo *schedulercache.NodeInfo) (schedulerapi.HostPriority, error) {
node := nodeInfo.Node()
if node == nil {
return schedulerapi.HostPriority{}, fmt.Errorf("node not found")
}
var sumSize int64
for i := range pod.Spec.Containers {
sumSize += checkContainerImageOnNode(node, &pod.Spec.Containers[i])
}
return schedulerapi.HostPriority{
Host: node.Name,
Score: calculateScoreFromSize(sumSize),
}, nil
}
// CheckPredicates Checks if the given pod can be placed on the given node.
func (p *PredicateChecker) CheckPredicates(pod *kube_api.Pod, nodeInfo *schedulercache.NodeInfo) error {
// TODO(fgrzadkowski): Use full list of predicates.
match, err := predicates.GeneralPredicates(pod, nodeInfo)
nodename := "unknown"
if nodeInfo.Node() != nil {
nodename = nodeInfo.Node().Name
}
if err != nil {
return fmt.Errorf("cannot put %s on %s due to %v", pod.Name, nodename, err)
}
if !match {
return fmt.Errorf("cannot put %s on %s", pod.Name, nodename)
}
return nil
}
func PodFitsResources(pod *api.Pod, meta interface{}, nodeInfo *schedulercache.NodeInfo) (bool, error) {
node := nodeInfo.Node()
if node == nil {
return false, fmt.Errorf("node not found")
}
allowedPodNumber := nodeInfo.AllowedPodNumber()
if len(nodeInfo.Pods())+1 > allowedPodNumber {
return false,
newInsufficientResourceError(podCountResourceName, 1, int64(len(nodeInfo.Pods())), int64(allowedPodNumber))
}
var podRequest *resourceRequest
predicateMeta, ok := meta.(*predicateMetadata)
if ok {
podRequest = predicateMeta.podRequest
} else {
// We couldn't parse metadata - fallback to computing it.
podRequest = getResourceRequest(pod)
}
if podRequest.milliCPU == 0 && podRequest.memory == 0 && podRequest.nvidiaGPU == 0 {
return true, nil
}
allocatable := node.Status.Allocatable
totalMilliCPU := allocatable.Cpu().MilliValue()
totalMemory := allocatable.Memory().Value()
totalNvidiaGPU := allocatable.NvidiaGPU().Value()
if totalMilliCPU < podRequest.milliCPU+nodeInfo.RequestedResource().MilliCPU {
return false,
newInsufficientResourceError(cpuResourceName, podRequest.milliCPU, nodeInfo.RequestedResource().MilliCPU, totalMilliCPU)
}
if totalMemory < podRequest.memory+nodeInfo.RequestedResource().Memory {
return false,
newInsufficientResourceError(memoryResourceName, podRequest.memory, nodeInfo.RequestedResource().Memory, totalMemory)
}
if totalNvidiaGPU < podRequest.nvidiaGPU+nodeInfo.RequestedResource().NvidiaGPU {
return false,
newInsufficientResourceError(nvidiaGpuResourceName, podRequest.nvidiaGPU, nodeInfo.RequestedResource().NvidiaGPU, totalNvidiaGPU)
}
if glog.V(10) {
// We explicitly don't do glog.V(10).Infof() to avoid computing all the parameters if this is
// not logged. There is visible performance gain from it.
glog.Infof("Schedule Pod %+v on Node %+v is allowed, Node is running only %v out of %v Pods.",
podName(pod), node.Name, len(nodeInfo.Pods()), allowedPodNumber)
}
return true, nil
}
// CheckNodeLabelPresence checks whether all of the specified labels exists on a node or not, regardless of their value
// If "presence" is false, then returns false if any of the requested labels matches any of the node's labels,
// otherwise returns true.
// If "presence" is true, then returns false if any of the requested labels does not match any of the node's labels,
// otherwise returns true.
//
// Consider the cases where the nodes are placed in regions/zones/racks and these are identified by labels
// In some cases, it is required that only nodes that are part of ANY of the defined regions/zones/racks be selected
//
// Alternately, eliminating nodes that have a certain label, regardless of value, is also useful
// A node may have a label with "retiring" as key and the date as the value
// and it may be desirable to avoid scheduling new pods on this node
func (n *NodeLabelChecker) CheckNodeLabelPresence(pod *api.Pod, meta interface{}, nodeInfo *schedulercache.NodeInfo) (bool, error) {
node := nodeInfo.Node()
if node == nil {
return false, fmt.Errorf("node not found")
}
var exists bool
nodeLabels := labels.Set(node.Labels)
for _, label := range n.labels {
exists = nodeLabels.Has(label)
if (exists && !n.presence) || (!exists && n.presence) {
return false, ErrNodeLabelPresenceViolated
}
}
return true, nil
}
// CheckPredicates checks if the given pod can be placed on the given node.
func (p *PredicateChecker) CheckPredicates(pod *kube_api.Pod, nodeInfo *schedulercache.NodeInfo) error {
for _, predicate := range p.predicates {
match, err := predicate(pod, nodeInfo)
nodename := "unknown"
if nodeInfo.Node() != nil {
nodename = nodeInfo.Node().Name
}
if err != nil {
return fmt.Errorf("cannot put %s on %s due to %v", pod.Name, nodename, err)
}
if !match {
return fmt.Errorf("cannot put %s on %s", pod.Name, nodename)
}
}
return nil
}
// CalculateNodeLabelPriority checks whether a particular label exists on a node or not, regardless of its value.
// If presence is true, prioritizes nodes that have the specified label, regardless of value.
// If presence is false, prioritizes nodes that do not have the specified label.
func (n *NodeLabelPrioritizer) CalculateNodeLabelPriorityMap(pod *api.Pod, meta interface{}, nodeInfo *schedulercache.NodeInfo) (schedulerapi.HostPriority, error) {
node := nodeInfo.Node()
if node == nil {
return schedulerapi.HostPriority{}, fmt.Errorf("node not found")
}
exists := labels.Set(node.Labels).Has(n.label)
score := 0
if (exists && n.presence) || (!exists && !n.presence) {
score = 10
}
return schedulerapi.HostPriority{
Host: node.Name,
Score: score,
}, nil
}
func PodFitsResources(pod *api.Pod, meta interface{}, nodeInfo *schedulercache.NodeInfo) (bool, []algorithm.PredicateFailureReason, error) {
node := nodeInfo.Node()
if node == nil {
return false, nil, fmt.Errorf("node not found")
}
var predicateFails []algorithm.PredicateFailureReason
allowedPodNumber := nodeInfo.AllowedPodNumber()
if len(nodeInfo.Pods())+1 > allowedPodNumber {
predicateFails = append(predicateFails, NewInsufficientResourceError(api.ResourcePods, 1, int64(len(nodeInfo.Pods())), int64(allowedPodNumber)))
}
var podRequest *schedulercache.Resource
if predicateMeta, ok := meta.(*predicateMetadata); ok {
podRequest = predicateMeta.podRequest
} else {
// We couldn't parse metadata - fallback to computing it.
podRequest = GetResourceRequest(pod)
}
if podRequest.MilliCPU == 0 && podRequest.Memory == 0 && podRequest.NvidiaGPU == 0 && len(podRequest.OpaqueIntResources) == 0 {
return len(predicateFails) == 0, predicateFails, nil
}
allocatable := nodeInfo.AllocatableResource()
if allocatable.MilliCPU < podRequest.MilliCPU+nodeInfo.RequestedResource().MilliCPU {
predicateFails = append(predicateFails, NewInsufficientResourceError(api.ResourceCPU, podRequest.MilliCPU, nodeInfo.RequestedResource().MilliCPU, allocatable.MilliCPU))
}
if allocatable.Memory < podRequest.Memory+nodeInfo.RequestedResource().Memory {
predicateFails = append(predicateFails, NewInsufficientResourceError(api.ResourceMemory, podRequest.Memory, nodeInfo.RequestedResource().Memory, allocatable.Memory))
}
if allocatable.NvidiaGPU < podRequest.NvidiaGPU+nodeInfo.RequestedResource().NvidiaGPU {
predicateFails = append(predicateFails, NewInsufficientResourceError(api.ResourceNvidiaGPU, podRequest.NvidiaGPU, nodeInfo.RequestedResource().NvidiaGPU, allocatable.NvidiaGPU))
}
for rName, rQuant := range podRequest.OpaqueIntResources {
if allocatable.OpaqueIntResources[rName] < rQuant+nodeInfo.RequestedResource().OpaqueIntResources[rName] {
predicateFails = append(predicateFails, NewInsufficientResourceError(rName, podRequest.OpaqueIntResources[rName], nodeInfo.RequestedResource().OpaqueIntResources[rName], allocatable.OpaqueIntResources[rName]))
}
}
if glog.V(10) {
// We explicitly don't do glog.V(10).Infof() to avoid computing all the parameters if this is
// not logged. There is visible performance gain from it.
glog.Infof("Schedule Pod %+v on Node %+v is allowed, Node is running only %v out of %v Pods.",
podName(pod), node.Name, len(nodeInfo.Pods()), allowedPodNumber)
}
return len(predicateFails) == 0, predicateFails, nil
}
// CalculateNodeAffinityPriority prioritizes nodes according to node affinity scheduling preferences
// indicated in PreferredDuringSchedulingIgnoredDuringExecution. Each time a node match a preferredSchedulingTerm,
// it will a get an add of preferredSchedulingTerm.Weight. Thus, the more preferredSchedulingTerms
// the node satisfies and the more the preferredSchedulingTerm that is satisfied weights, the higher
// score the node gets.
func CalculateNodeAffinityPriorityMap(pod *api.Pod, meta interface{}, nodeInfo *schedulercache.NodeInfo) (schedulerapi.HostPriority, error) {
node := nodeInfo.Node()
if node == nil {
return schedulerapi.HostPriority{}, fmt.Errorf("node not found")
}
var affinity *api.Affinity
if priorityMeta, ok := meta.(*priorityMetadata); ok {
affinity = priorityMeta.affinity
} else {
// We couldn't parse metadata - fallback to computing it.
var err error
affinity, err = api.GetAffinityFromPodAnnotations(pod.Annotations)
if err != nil {
return schedulerapi.HostPriority{}, err
}
}
var count int32
// A nil element of PreferredDuringSchedulingIgnoredDuringExecution matches no objects.
// An element of PreferredDuringSchedulingIgnoredDuringExecution that refers to an
// empty PreferredSchedulingTerm matches all objects.
if affinity != nil && affinity.NodeAffinity != nil && affinity.NodeAffinity.PreferredDuringSchedulingIgnoredDuringExecution != nil {
// Match PreferredDuringSchedulingIgnoredDuringExecution term by term.
for i := range affinity.NodeAffinity.PreferredDuringSchedulingIgnoredDuringExecution {
preferredSchedulingTerm := &affinity.NodeAffinity.PreferredDuringSchedulingIgnoredDuringExecution[i]
if preferredSchedulingTerm.Weight == 0 {
continue
}
// TODO: Avoid computing it for all nodes if this becomes a performance problem.
nodeSelector, err := api.NodeSelectorRequirementsAsSelector(preferredSchedulingTerm.Preference.MatchExpressions)
if err != nil {
return schedulerapi.HostPriority{}, err
}
if nodeSelector.Matches(labels.Set(node.Labels)) {
count += preferredSchedulingTerm.Weight
}
}
}
return schedulerapi.HostPriority{
Host: node.Name,
Score: int(count),
}, nil
}
func (t *TolerationMatch) PodToleratesNodeTaints(pod *api.Pod, nodeInfo *schedulercache.NodeInfo) (bool, error) {
node := nodeInfo.Node()
taints, err := api.GetTaintsFromNodeAnnotations(node.Annotations)
if err != nil {
return false, err
}
tolerations, err := api.GetTolerationsFromPodAnnotations(pod.Annotations)
if err != nil {
return false, err
}
if tolerationsToleratesTaints(tolerations, taints) {
return true, nil
}
return false, ErrTaintsTolerationsNotMatch
}
func calculateBalancedResourceAllocation(pod *api.Pod, podRequests *schedulercache.Resource, nodeInfo *schedulercache.NodeInfo) (schedulerapi.HostPriority, error) {
node := nodeInfo.Node()
if node == nil {
return schedulerapi.HostPriority{}, fmt.Errorf("node not found")
}
allocatableResources := nodeInfo.AllocatableResource()
totalResources := *podRequests
totalResources.MilliCPU += nodeInfo.NonZeroRequest().MilliCPU
totalResources.Memory += nodeInfo.NonZeroRequest().Memory
cpuFraction := fractionOfCapacity(totalResources.MilliCPU, allocatableResources.MilliCPU)
memoryFraction := fractionOfCapacity(totalResources.Memory, allocatableResources.Memory)
score := int(0)
if cpuFraction >= 1 || memoryFraction >= 1 {
// if requested >= capacity, the corresponding host should never be preferred.
score = 0
} else {
// Upper and lower boundary of difference between cpuFraction and memoryFraction are -1 and 1
// respectively. Multilying the absolute value of the difference by 10 scales the value to
// 0-10 with 0 representing well balanced allocation and 10 poorly balanced. Subtracting it from
// 10 leads to the score which also scales from 0 to 10 while 10 representing well balanced.
diff := math.Abs(cpuFraction - memoryFraction)
score = int(10 - diff*10)
}
if glog.V(10) {
// We explicitly don't do glog.V(10).Infof() to avoid computing all the parameters if this is
// not logged. There is visible performance gain from it.
glog.V(10).Infof(
"%v -> %v: Balanced Resource Allocation, capacity %d millicores %d memory bytes, total request %d millicores %d memory bytes, score %d",
pod.Name, node.Name,
allocatableResources.MilliCPU, allocatableResources.Memory,
totalResources.MilliCPU, totalResources.Memory,
score,
)
}
return schedulerapi.HostPriority{
Host: node.Name,
Score: score,
}, nil
}
// CheckNodeMemoryPressurePredicate checks if a pod can be scheduled on a node
// reporting memory pressure condition.
func CheckNodeMemoryPressurePredicate(pod *api.Pod, nodeInfo *schedulercache.NodeInfo) (bool, error) {
node := nodeInfo.Node()
if node == nil {
return false, fmt.Errorf("node not found")
}
// pod is not BestEffort pod
if !isPodBestEffort(pod) {
return true, nil
}
// is node under presure?
for _, cond := range node.Status.Conditions {
if cond.Type == api.NodeMemoryPressure && cond.Status == api.ConditionTrue {
return false, ErrNodeUnderMemoryPressure
}
}
return true, nil
}
func PodToleratesNodeTaints(pod *v1.Pod, meta interface{}, nodeInfo *schedulercache.NodeInfo) (bool, []algorithm.PredicateFailureReason, error) {
node := nodeInfo.Node()
if node == nil {
return false, nil, fmt.Errorf("node not found")
}
taints, err := v1.GetTaintsFromNodeAnnotations(node.Annotations)
if err != nil {
return false, nil, err
}
tolerations, err := v1.GetTolerationsFromPodAnnotations(pod.Annotations)
if err != nil {
return false, nil, err
}
if tolerationsToleratesTaints(tolerations, taints) {
return true, nil, nil
}
return false, []algorithm.PredicateFailureReason{ErrTaintsTolerationsNotMatch}, nil
}
func PodToleratesNodeTaints(pod *api.Pod, meta interface{}, nodeInfo *schedulercache.NodeInfo) (bool, error) {
node := nodeInfo.Node()
if node == nil {
return false, fmt.Errorf("node not found")
}
taints, err := api.GetTaintsFromNodeAnnotations(node.Annotations)
if err != nil {
return false, err
}
tolerations, err := api.GetTolerationsFromPodAnnotations(pod.Annotations)
if err != nil {
return false, err
}
if tolerationsToleratesTaints(tolerations, taints) {
return true, nil
}
return false, ErrTaintsTolerationsNotMatch
}