image.png

image.png

PodAffinity

所有PodAffinityTerm都必须满足

  1. // Pod affinity is a group of inter pod affinity scheduling rules.
  2. type PodAffinity struct {
  3.     // NOT YET IMPLEMENTED. TODO: Uncomment field once it is implemented.
  4.     // If the affinity requirements specified by this field are not met at
  5.     // scheduling time, the pod will not be scheduled onto the node.
  6.     // If the affinity requirements specified by this field cease to be met
  7.     // at some point during pod execution (e.g. due to a pod label update), the
  8.     // system will try to eventually evict the pod from its node.
  9.     // When there are multiple elements, the lists of nodes corresponding to each
  10.     // podAffinityTerm are intersected, i.e. all terms must be satisfied.
  11.     // +optional
  12.     // RequiredDuringSchedulingRequiredDuringExecution []PodAffinityTerm  `json:"requiredDuringSchedulingRequiredDuringExecution,omitempty"`
  14.     // If the affinity requirements specified by this field are not met at
  15.     // scheduling time, the pod will not be scheduled onto the node.
  16.     // If the affinity requirements specified by this field cease to be met
  17.     // at some point during pod execution (e.g. due to a pod label update), the
  18.     // system may or may not try to eventually evict the pod from its node.
  19.     // When there are multiple elements, the lists of nodes corresponding to each
  20.     // podAffinityTerm are intersected, i.e. all terms must be satisfied.
  21.     // +optional
  22.     RequiredDuringSchedulingIgnoredDuringExecution []PodAffinityTerm `json:"requiredDuringSchedulingIgnoredDuringExecution,omitempty" protobuf:"bytes,1,rep,name=requiredDuringSchedulingIgnoredDuringExecution"`
  23.     // The scheduler will prefer to schedule pods to nodes that satisfy
  24.     // the affinity expressions specified by this field, but it may choose
  25.     // a node that violates one or more of the expressions. The node that is
  26.     // most preferred is the one with the greatest sum of weights, i.e.
  27.     // for each node that meets all of the scheduling requirements (resource
  28.     // request, requiredDuringScheduling affinity expressions, etc.),
  29.     // compute a sum by iterating through the elements of this field and adding
  30.     // "weight" to the sum if the node has pods which matches the corresponding podAffinityTerm; the
  31.     // node(s) with the highest sum are the most preferred.
  32.     // +optional
  33.     PreferredDuringSchedulingIgnoredDuringExecution []WeightedPodAffinityTerm `json:"preferredDuringSchedulingIgnoredDuringExecution,omitempty" protobuf:"bytes,2,rep,name=preferredDuringSchedulingIgnoredDuringExecution"`
  34. }

PodAffinityTerm

一个PodAffinity包含多个PodAffinityTerm，多个PodAffinityTerm是AND关系。
一个PodAffinityTerm对应一个LabelSelector

  1. // Defines a set of pods (namely those matching the labelSelector
  2. // relative to the given namespace(s)) that this pod should be
  3. // co-located (affinity) or not co-located (anti-affinity) with,
  4. // where co-located is defined as running on a node whose value of
  5. // the label with key <topologyKey> matches that of any node on which
  6. // a pod of the set of pods is running
  7. type PodAffinityTerm struct {
  8.     // A label query over a set of resources, in this case pods.
  9.     // +optional
  10.     LabelSelector *metav1.LabelSelector `json:"labelSelector,omitempty" protobuf:"bytes,1,opt,name=labelSelector"`
  11.     // namespaces specifies which namespaces the labelSelector applies to (matches against);
  12.     // null or empty list means "this pod's namespace"
  13.     // +optional
  14.     Namespaces []string `json:"namespaces,omitempty" protobuf:"bytes,2,rep,name=namespaces"`
  15.     // This pod should be co-located (affinity) or not co-located (anti-affinity) with the pods matching
  16.     // the labelSelector in the specified namespaces, where co-located is defined as running on a node
  17.     // whose value of the label with key topologyKey matches that of any node on which any of the
  18.     // selected pods is running.
  19.     // Empty topologyKey is not allowed.
  20.     TopologyKey string `json:"topologyKey" protobuf:"bytes,3,opt,name=topologyKey"`
  21. }

LabelSelector

matchLabels 是由 {key,value} 对组成的映射。 matchLabels 映射中的单个 {key,value } 等同于 matchExpressions 的元素， 其 key 字段为 “key”，operator 为 “In”，而 values 数组仅包含 “value”。

matchExpressions 是 Pod 选择算符需求的列表。 有效的运算符包括 In、NotIn、Exists 和 DoesNotExist。 在 In 和 NotIn 的情况下，设置的值必须是非空的。 matchExpressions是数组：[]LabelSelectorRequirement，多个LabelSelectorRequirement之间是AND关系

来自 matchLabels 和 matchExpressions 的所有要求都按逻辑与的关系组合到一起 — 它们必须都满足才能匹配。**

  1. // Note:
  2. // There are two different styles of label selectors used in versioned types:
  3. // an older style which is represented as just a string in versioned types, and a
  4. // newer style that is structured.  LabelSelector is an internal representation for the
  5. // latter style.
  7. // A label selector is a label query over a set of resources. The result of matchLabels and
  8. // matchExpressions are ANDed. An empty label selector matches all objects. A null
  9. // label selector matches no objects.
  10. type LabelSelector struct {
  11.     // matchLabels is a map of {key,value} pairs. A single {key,value} in the matchLabels
  12.     // map is equivalent to an element of matchExpressions, whose key field is "key", the
  13.     // operator is "In", and the values array contains only "value". The requirements are ANDed.
  14.     // +optional
  15.     MatchLabels map[string]string `json:"matchLabels,omitempty" protobuf:"bytes,1,rep,name=matchLabels"`
  16.     // matchExpressions is a list of label selector requirements. The requirements are ANDed.
  17.     // +optional
  18.     MatchExpressions []LabelSelectorRequirement `json:"matchExpressions,omitempty" protobuf:"bytes,2,rep,name=matchExpressions"`
  19. }

NodeAffinity

  1. // Node affinity is a group of node affinity scheduling rules.
  2. type NodeAffinity struct {
  3.     // NOT YET IMPLEMENTED. TODO: Uncomment field once it is implemented.
  4.     // If the affinity requirements specified by this field are not met at
  5.     // scheduling time, the pod will not be scheduled onto the node.
  6.     // If the affinity requirements specified by this field cease to be met
  7.     // at some point during pod execution (e.g. due to an update), the system
  8.     // will try to eventually evict the pod from its node.
  9.     // +optional
  10.     // RequiredDuringSchedulingRequiredDuringExecution *NodeSelector `json:"requiredDuringSchedulingRequiredDuringExecution,omitempty"`
  12.     // If the affinity requirements specified by this field are not met at
  13.     // scheduling time, the pod will not be scheduled onto the node.
  14.     // If the affinity requirements specified by this field cease to be met
  15.     // at some point during pod execution (e.g. due to an update), the system
  16.     // may or may not try to eventually evict the pod from its node.
  17.     // +optional
  18.     RequiredDuringSchedulingIgnoredDuringExecution *NodeSelector `json:"requiredDuringSchedulingIgnoredDuringExecution,omitempty" protobuf:"bytes,1,opt,name=requiredDuringSchedulingIgnoredDuringExecution"`
  19.     // The scheduler will prefer to schedule pods to nodes that satisfy
  20.     // the affinity expressions specified by this field, but it may choose
  21.     // a node that violates one or more of the expressions. The node that is
  22.     // most preferred is the one with the greatest sum of weights, i.e.
  23.     // for each node that meets all of the scheduling requirements (resource
  24.     // request, requiredDuringScheduling affinity expressions, etc.),
  25.     // compute a sum by iterating through the elements of this field and adding
  26.     // "weight" to the sum if the node matches the corresponding matchExpressions; the
  27.     // node(s) with the highest sum are the most preferred.
  28.     // +optional
  29.     PreferredDuringSchedulingIgnoredDuringExecution []PreferredSchedulingTerm `json:"preferredDuringSchedulingIgnoredDuringExecution,omitempty" protobuf:"bytes,2,rep,name=preferredDuringSchedulingIgnoredDuringExecution"`
  30. }

NodeSelector

一个NodeAffinity对一个NodeSelector。
一个NodeSelector可以配置多个NodeSelectorTerm，多个NodeSelectorTerm之间是OR关系

  1. // A node selector represents the union of the results of one or more label queries
  2. // over a set of nodes; that is, it represents the OR of the selectors represented
  3. // by the node selector terms.
  4. type NodeSelector struct {
  5.     //Required. A list of node selector terms. The terms are ORed.
  6.     NodeSelectorTerms []NodeSelectorTerm `json:"nodeSelectorTerms" protobuf:"bytes,1,rep,name=nodeSelectorTerms"`
  7. }

NodeSelectorTerm

每个NodeSelectorTerm有一个属性matchExpressions，matchExpression数组是AND关系

  1. // A null or empty node selector term matches no objects. The requirements of
  2. // them are ANDed.
  3. // The TopologySelectorTerm type implements a subset of the NodeSelectorTerm.
  4. type NodeSelectorTerm struct {
  5.     // A list of node selector requirements by node's labels.
  6.     // +optional
  7.     MatchExpressions []NodeSelectorRequirement `json:"matchExpressions,omitempty" protobuf:"bytes,1,rep,name=matchExpressions"`
  8.     // A list of node selector requirements by node's fields.
  9.     // +optional
  10.     MatchFields []NodeSelectorRequirement `json:"matchFields,omitempty" protobuf:"bytes,2,rep,name=matchFields"`
  11. }