diff --git a/pkg/controller/network_policy.go b/pkg/controller/network_policy.go
index 2a7a62cc966..f7b54c423a6 100644
--- a/pkg/controller/network_policy.go
+++ b/pkg/controller/network_policy.go
@@ -152,7 +152,7 @@ func (c *Controller) handleUpdateNp(key string) error {
 		return err
 	}
 
-	if err = c.checkAndUpdateNodePortGroup(false); err != nil {
+	if err = c.checkAndUpdateNodePortGroup(false, ""); err != nil {
 		klog.Errorf("failed to update node acl: %v", err)
 		return err
 	}
@@ -571,6 +571,11 @@ func (c *Controller) handleDeleteNp(key string) error {
 	defer func() { _ = c.npKeyMutex.UnlockKey(key) }()
 	klog.Infof("handle delete network policy %s", key)
 
+	if err = c.checkAndUpdateNodePortGroup(false, ""); err != nil {
+		klog.Errorf("failed to update node acl: %v", err)
+		return err
+	}
+
 	npName := name
 	nameArray := []rune(name)
 	if !unicode.IsLetter(nameArray[0]) {
diff --git a/pkg/controller/node.go b/pkg/controller/node.go
index ef0d77fc190..d2d96cdaf6e 100644
--- a/pkg/controller/node.go
+++ b/pkg/controller/node.go
@@ -933,14 +933,18 @@ func (c *Controller) fetchPodsOnNode(nodeName string, pods []*v1.Pod) ([]string,
 }
 
 func (c *Controller) CheckNodePortGroup() {
-	if err := c.checkAndUpdateNodePortGroup(true); err != nil {
+	if err := c.checkAndUpdateNodePortGroup(true, ""); err != nil {
 		klog.Errorf("check node port group status: %v", err)
 	}
 }
 
 var nodeAclExists bool
 
-func (c *Controller) checkAndUpdateNodePortGroup(updateIfNotExists bool) error {
+func (c *Controller) checkAndUpdateNodePortGroup(alwaysUpdate bool, nodeName string) error {
+	if !c.config.EnableNP {
+		return nil
+	}
+
 	c.npKeyMutex.LockKey("node_acl")
 	defer func() { _ = c.npKeyMutex.UnlockKey("node_acl") }()
 
@@ -948,19 +952,29 @@ func (c *Controller) checkAndUpdateNodePortGroup(updateIfNotExists bool) error {
 	np, _ := c.npsLister.List(labels.Everything())
 	networkPolicyExists := len(np) != 0
 
-	if !updateIfNotExists && networkPolicyExists == nodeAclExists {
+	if !alwaysUpdate && networkPolicyExists == nodeAclExists {
 		return nil
 	}
 
-	nodes, err := c.nodesLister.List(labels.Everything())
+	pods, err := c.podsLister.List(labels.Everything())
 	if err != nil {
-		klog.Errorf("list nodes: %v", err)
+		klog.Errorf("failed to list pods: %v", err)
 		return err
 	}
 
-	pods, err := c.podsLister.List(labels.Everything())
-	if err != nil {
-		klog.Errorf("list pods, %v", err)
+	var nodes []*v1.Node
+	if nodeName != "" {
+		node, err := c.nodesLister.Get(nodeName)
+		if err != nil {
+			if k8serrors.IsNotFound(err) {
+				return nil
+			}
+			klog.Errorf("failed to get node %q: %v", err)
+			return err
+		}
+		nodes = []*v1.Node{node}
+	} else if nodes, err = c.nodesLister.List(labels.Everything()); err != nil {
+		klog.Errorf("failed to list nodes: %v", err)
 		return err
 	}
 
@@ -1003,7 +1017,10 @@ func (c *Controller) checkAndUpdateNodePortGroup(updateIfNotExists bool) error {
 		}
 	}
 
-	nodeAclExists = networkPolicyExists
+	if nodeName == "" {
+		nodeAclExists = networkPolicyExists
+	}
+
 	return nil
 }
 
diff --git a/pkg/controller/pod.go b/pkg/controller/pod.go
index 98dcda78b7e..22f53eda4d9 100644
--- a/pkg/controller/pod.go
+++ b/pkg/controller/pod.go
@@ -179,8 +179,12 @@ func (c *Controller) enqueueAddPod(obj interface{}) {
 	}
 
 	p := obj.(*v1.Pod)
+	workqueueKey := key
 	// TODO: we need to find a way to reduce duplicated np added to the queue
 	if c.config.EnableNP {
+		if p.Spec.NodeName != "" {
+			workqueueKey = fmt.Sprintf("%s@%s", key, p.Spec.NodeName)
+		}
 		c.namedPort.AddNamedPortByPod(p)
 		if p.Status.PodIP != "" {
 			for _, np := range c.podMatchNetworkPolicies(p) {
@@ -222,7 +226,7 @@ func (c *Controller) enqueueAddPod(obj interface{}) {
 	}
 	if need {
 		klog.Infof("enqueue add pod %s", key)
-		c.addOrUpdatePodQueue.Add(key)
+		c.addOrUpdatePodQueue.Add(workqueueKey)
 	}
 }
 
@@ -270,7 +274,12 @@ func (c *Controller) enqueueUpdatePod(oldObj, newObj interface{}) {
 		return
 	}
 
+	workqueueKey := key
 	if c.config.EnableNP {
+		if newPod.Spec.NodeName != oldPod.Spec.NodeName {
+			workqueueKey = fmt.Sprintf("%s@%s", key, newPod.Spec.NodeName)
+		}
+
 		c.namedPort.AddNamedPortByPod(newPod)
 		newNp := c.podMatchNetworkPolicies(newPod)
 		if !reflect.DeepEqual(oldPod.Labels, newPod.Labels) {
@@ -345,7 +354,7 @@ func (c *Controller) enqueueUpdatePod(oldObj, newObj interface{}) {
 		return
 	}
 	klog.Infof("enqueue update pod %s", key)
-	c.addOrUpdatePodQueue.Add(key)
+	c.addOrUpdatePodQueue.Add(workqueueKey)
 
 	// security policy changed
 	for _, podNet := range podNets {
@@ -544,15 +553,20 @@ func (c *Controller) changeVMSubnet(vmName, namespace, providerName, subnetName
 }
 
 func (c *Controller) handleAddOrUpdatePod(key string) (err error) {
-	namespace, name, err := cache.SplitMetaNamespaceKey(key)
+	podKey := key
+	if idx := strings.IndexRune(key, '@'); idx >= 0 {
+		podKey = key[:idx]
+	}
+
+	namespace, name, err := cache.SplitMetaNamespaceKey(podKey)
 	if err != nil {
-		utilruntime.HandleError(fmt.Errorf("invalid resource key: %s", key))
+		utilruntime.HandleError(fmt.Errorf("invalid resource key: %s", podKey))
 		return nil
 	}
 
-	c.podKeyMutex.LockKey(key)
-	defer func() { _ = c.podKeyMutex.UnlockKey(key) }()
-	klog.Infof("handle add/update pod %s", key)
+	c.podKeyMutex.LockKey(podKey)
+	defer func() { _ = c.podKeyMutex.UnlockKey(podKey) }()
+	klog.Infof("handle add/update pod %s", podKey)
 
 	cachedPod, err := c.podsLister.Pods(namespace).Get(name)
 	if err != nil {
@@ -600,6 +614,13 @@ func (c *Controller) handleAddOrUpdatePod(key string) (err error) {
 		}
 	}
 
+	if len(podKey) != len(key) && pod.Spec.NodeName != "" {
+		if err = c.checkAndUpdateNodePortGroup(true, pod.Spec.NodeName); err != nil {
+			klog.Errorf("failed to sync node port group for pod %s on node %s: %v", podKey, pod.Spec.NodeName, err)
+			return err
+		}
+	}
+
 	// check if route subnet is need.
 	pod = cachedPod.DeepCopy()
 	return c.reconcileRouteSubnets(cachedPod, pod, needRouteSubnets(pod, podNets))