diff --git a/pkg/binder/utils/binderunit.go b/pkg/binder/utils/binderunit.go
index ca935347..dca1a546 100644
--- a/pkg/binder/utils/binderunit.go
+++ b/pkg/binder/utils/binderunit.go
@@ -70,6 +70,7 @@ func GetUnitType(pod *v1.Pod) framework.ScheduleUnitType {
 // In this case, we won't move pods from activeQ to readyQ, caller should handle the not found issue.
 func CreateScheduleUnit(pcLister schedulingv1.PriorityClassLister, pgLister v1alpha1.PodGroupLister, info *framework.QueuedPodInfo) (framework.ScheduleUnit, error) {
 	if len(unitutil.GetPodGroupName(info.Pod)) != 0 {
+		/// TODO
 		pgName := unitutil.GetPodGroupName(info.Pod)
 		podGroup, err := pgLister.PodGroups(info.Pod.Namespace).Get(pgName)
 		if err != nil {
diff --git a/pkg/scheduler/core/pod_scheduler/pod_scheduler_test.go b/pkg/scheduler/core/pod_scheduler/pod_scheduler_test.go
index e96429c3..329e906f 100644
--- a/pkg/scheduler/core/pod_scheduler/pod_scheduler_test.go
+++ b/pkg/scheduler/core/pod_scheduler/pod_scheduler_test.go
@@ -50,12 +50,11 @@ func TestSelectHost(t *testing.T) {
 		isolatedCache: isolatedcache.NewIsolatedCache(),
 	}
 	tests := []struct {
-		name               string
-		list               framework.NodeScoreList
-		possibleHosts      sets.String
-		expectsErr         bool
-		pod                *v1.Pod
-		expectedCacheSizes []int
+		name          string
+		list          framework.NodeScoreList
+		possibleHosts sets.String
+		expectsErr    bool
+		pod           *v1.Pod
 	}{
 		{
 			name: "unique properly ordered scores",
@@ -67,7 +66,6 @@ func TestSelectHost(t *testing.T) {
 			expectsErr:    false,
 			pod: testinghelper.MakePod().Namespace("default").Name("pod1").UID("pod1").
 				ControllerRef(metav1.OwnerReference{Kind: "ReplicaSet", Name: "rs1", UID: "rs1"}).Obj(),
-			expectedCacheSizes: []int{1, 2, 3, 4, 5, 6, 7, 8, 9, 10},
 		},
 		{
 			name: "equal scores",
@@ -81,7 +79,6 @@ func TestSelectHost(t *testing.T) {
 			expectsErr:    false,
 			pod: testinghelper.MakePod().Namespace("default").Name("pod2").UID("pod2").
 				ControllerRef(metav1.OwnerReference{Kind: "ReplicaSet", Name: "rs2", UID: "rs2"}).Obj(),
-			expectedCacheSizes: []int{3, 6, 9, 12, 15, 18, 21, 24, 27, 30},
 		},
 		{
 			name: "out of order scores",
@@ -96,7 +93,6 @@ func TestSelectHost(t *testing.T) {
 			expectsErr:    false,
 			pod: testinghelper.MakePod().Namespace("default").Name("pod3").UID("pod3").
 				ControllerRef(metav1.OwnerReference{Kind: "StatefulSet", Name: "sts3", UID: "sts3"}).Obj(),
-			expectedCacheSizes: []int{4, 8, 12, 16, 20, 24, 28, 32, 36, 40},
 		},
 		{
 			name:          "empty priority list",
@@ -105,34 +101,38 @@ func TestSelectHost(t *testing.T) {
 			expectsErr:    true,
 			pod: testinghelper.MakePod().Namespace("default").Name("pod4").UID("pod4").
 				ControllerRef(metav1.OwnerReference{Kind: "StatefulSet", Name: "sts4", UID: "sts4"}).Obj(),
-			expectedCacheSizes: []int{0, 0, 0, 0, 0, 0, 0, 0, 0, 0},
 		},
 	}
 
 	for _, test := range tests {
 		t.Run(test.name, func(t *testing.T) {
-			lastCacheSize := 0
 			// get mock pod owner
 			podOwner := podutil.GetPodOwner(test.pod)
 			// increase the randomness
 			for i := 0; i < 10; i++ {
 				// TODO: add unit test cases for caching node logic
+				// init cache
+				nonSelectedNodes := sets.NewString()
+				for _, nodeScore := range test.list {
+					nodeName := nodeScore.Name
+					scheduler.isolatedCache.DeleteNodeForPodOwner(podOwner, nodeName)
+					nonSelectedNodes.Insert(nodeName)
+				}
 
 				selectedNode, err := scheduler.selectHostAndCacheResults(test.list, &v1.Pod{}, podOwner, "", &framework.UnitSchedulingRequest{})
 				nodes := scheduler.isolatedCache.GetOrderedNodesForPodOwner(podOwner)
 
-				currCacheSize := len(nodes)
-				if currCacheSize != test.expectedCacheSizes[i] {
-					t.Errorf("Unexpected cache size, expected %d but got %d", test.expectedCacheSizes[i], currCacheSize)
-				}
-				lastCacheSize = currCacheSize
+				// exclude the selected node
+				nonSelectedNodes.Delete(selectedNode)
 
-				// Check newly cached nodes
-				newlyCachedNodes := nodes[lastCacheSize:]
-				for _, newlyCachedNode := range newlyCachedNodes {
-					if selectedNode == newlyCachedNode {
+				for _, node := range nodes {
+					if selectedNode == node {
 						t.Errorf("Unexpected cached node %s, selected node should not be cached", selectedNode)
 					}
+					if !nonSelectedNodes.Has(node) {
+						t.Errorf("Non-selected nodes should be cached")
+					}
+					nonSelectedNodes.Delete(node)
 				}
 
 				if test.expectsErr {