Merge pull request #60386 from k82cn/k8s_60163

Automatic merge from submit-queue (batch tested with PRs 60683, 60386). If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.

Added unschedulabe predicate.

Signed-off-by: Da K. Ma <madaxa@cn.ibm.com>

**Which issue(s) this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close the issue(s) when PR gets merged)*:
Fixes #60163

**Release note**:
```release-note
None
```

pkg/controller/daemon/daemon_controller_test.go

@@ -1545,6 +1545,7 @@ func TestNodeShouldRunDaemonPod(t *testing.T) {
 		predicateName                                    string
 		podsOnNode                                       []*v1.Pod
 		nodeCondition                                    []v1.NodeCondition
+		nodeUnschedulable                                bool
 		ds                                               *apps.DaemonSet
 		wantToRun, shouldSchedule, shouldContinueRunning bool
 		err                                              error
@@ -1800,6 +1801,24 @@ func TestNodeShouldRunDaemonPod(t *testing.T) {
 			shouldSchedule:        true,
 			shouldContinueRunning: true,
 		},
+		{
+			predicateName: "ShouldRunDaemonPodOnUnscheduableNode",
+			ds: &apps.DaemonSet{
+				Spec: apps.DaemonSetSpec{
+					Selector: &metav1.LabelSelector{MatchLabels: simpleDaemonSetLabel},
+					Template: v1.PodTemplateSpec{
+						ObjectMeta: metav1.ObjectMeta{
+							Labels: simpleDaemonSetLabel,
+						},
+						Spec: resourcePodSpec("", "50M", "0.5"),
+					},
+				},
+			},
+			nodeUnschedulable:     true,
+			wantToRun:             true,
+			shouldSchedule:        true,
+			shouldContinueRunning: true,
+		},
 	}
 
 	for i, c := range cases {
@@ -1807,6 +1826,7 @@ func TestNodeShouldRunDaemonPod(t *testing.T) {
 			node := newNode("test-node", simpleDaemonSetLabel)
 			node.Status.Conditions = append(node.Status.Conditions, c.nodeCondition...)
 			node.Status.Allocatable = allocatableResources("100M", "1")
+			node.Spec.Unschedulable = c.nodeUnschedulable
 			manager, _, _, err := newTestController()
 			if err != nil {
 				t.Fatalf("error creating DaemonSets controller: %v", err)

pkg/scheduler/algorithm/predicates/predicates.go

@@ -69,6 +69,8 @@ const (
 	NoDiskConflictPred = "NoDiskConflict"
 	// PodToleratesNodeTaintsPred defines the name of predicate PodToleratesNodeTaints.
 	PodToleratesNodeTaintsPred = "PodToleratesNodeTaints"
+	// CheckNodeUnschedulablePred defines the name of predicate CheckNodeUnschedulablePredicate.
+	CheckNodeUnschedulablePred = "CheckNodeUnschedulable"
 	// PodToleratesNodeNoExecuteTaintsPred defines the name of predicate PodToleratesNodeNoExecuteTaints.
 	PodToleratesNodeNoExecuteTaintsPred = "PodToleratesNodeNoExecuteTaints"
 	// CheckNodeLabelPresencePred defines the name of predicate CheckNodeLabelPresence.
@@ -125,7 +127,7 @@ const (
 // The order is based on the restrictiveness & complexity of predicates.
 // Design doc: https://github.com/kubernetes/community/blob/master/contributors/design-proposals/scheduling/predicates-ordering.md
 var (
-	predicatesOrdering = []string{CheckNodeConditionPred,
+	predicatesOrdering = []string{CheckNodeConditionPred, CheckNodeUnschedulablePred,
 		GeneralPred, HostNamePred, PodFitsHostPortsPred,
 		MatchNodeSelectorPred, PodFitsResourcesPred, NoDiskConflictPred,
 		PodToleratesNodeTaintsPred, PodToleratesNodeNoExecuteTaintsPred, CheckNodeLabelPresencePred,
@@ -1446,8 +1448,8 @@ func (c *PodAffinityChecker) satisfiesPodsAffinityAntiAffinity(pod *v1.Pod, node
 	return nil, nil
 }
 
-// PodToleratesNodeTaints checks if a pod tolerations can tolerate the node taints
-func PodToleratesNodeTaints(pod *v1.Pod, meta algorithm.PredicateMetadata, nodeInfo *schedulercache.NodeInfo) (bool, []algorithm.PredicateFailureReason, error) {
+// CheckNodeUnschedulablePredicate checks if a pod can be scheduled on a node with Unschedulable spec.
+func CheckNodeUnschedulablePredicate(pod *v1.Pod, meta algorithm.PredicateMetadata, nodeInfo *schedulercache.NodeInfo) (bool, []algorithm.PredicateFailureReason, error) {
 	if nodeInfo == nil || nodeInfo.Node() == nil {
 		return false, []algorithm.PredicateFailureReason{ErrNodeUnknownCondition}, nil
 	}
@@ -1456,6 +1458,15 @@ func PodToleratesNodeTaints(pod *v1.Pod, meta algorithm.PredicateMetadata, nodeI
 		return false, []algorithm.PredicateFailureReason{ErrNodeUnschedulable}, nil
 	}
 
+	return true, nil, nil
+}
+
+// PodToleratesNodeTaints checks if a pod tolerations can tolerate the node taints
+func PodToleratesNodeTaints(pod *v1.Pod, meta algorithm.PredicateMetadata, nodeInfo *schedulercache.NodeInfo) (bool, []algorithm.PredicateFailureReason, error) {
+	if nodeInfo == nil || nodeInfo.Node() == nil {
+		return false, []algorithm.PredicateFailureReason{ErrNodeUnknownCondition}, nil
+	}
+
 	return podToleratesNodeTaints(pod, nodeInfo, func(t *v1.Taint) bool {
 		// PodToleratesNodeTaints is only interested in NoSchedule and NoExecute taints.
 		return t.Effect == v1.TaintEffectNoSchedule || t.Effect == v1.TaintEffectNoExecute

pkg/scheduler/algorithmprovider/defaults/defaults.go

@@ -186,12 +186,15 @@ func ApplyFeatureGates() {
 		// if you just want remove specific provider, call func RemovePredicateKeyFromAlgoProvider()
 		factory.RemovePredicateKeyFromAlgorithmProviderMap(predicates.CheckNodeConditionPred)
 
+		// Fit is determined based on whether a node has Unschedulable spec
+		factory.RegisterMandatoryFitPredicate(predicates.CheckNodeUnschedulablePred, predicates.CheckNodeUnschedulablePredicate)
 		// Fit is determined based on whether a pod can tolerate all of the node's taints
 		factory.RegisterMandatoryFitPredicate(predicates.PodToleratesNodeTaintsPred, predicates.PodToleratesNodeTaints)
-		// Insert Key "PodToleratesNodeTaints" To All Algorithm Provider
+		// Insert Key "PodToleratesNodeTaints" and "CheckNodeUnschedulable" To All Algorithm Provider
 		// The key will insert to all providers which in algorithmProviderMap[]
 		// if you just want insert to specific provider, call func InsertPredicateKeyToAlgoProvider()
 		factory.InsertPredicateKeyToAlgorithmProviderMap(predicates.PodToleratesNodeTaintsPred)
+		factory.InsertPredicateKeyToAlgorithmProviderMap(predicates.CheckNodeUnschedulablePred)
 
 		glog.Warningf("TaintNodesByCondition is enabled, PodToleratesNodeTaints predicate is mandatory")
 	}