Kubernetes network polices are not enforced unless the network-plugin daemon-sets are restarted. Why?

I have only one network policy in my cluster in prod namespace that allows only ingress rules. The network plugin is weave-net. No rules are configured for Egress so I am expecting egress traffic will be blocked. But until I restart the network daemon-set pods the rule has no effect. I know by best practices I should have default ingress and egress rules. But I want to understand the reason of this behavior. Is this step always required to restart the network-plugin pods?

1. Network Policy Definition

apiVersion: networking.k8s.io/v1
kind: NetworkPolicy
metadata:
  name: test-network-policy
  namespace: prod
spec:
  ingress:
  - {}
  podSelector:
    matchLabels:
      run: prod-nginx
  policyTypes:
  - Ingress
  - Egress

2. Checking the netpol object

Name:         test-network-policy
Namespace:    prod
Created on:   2021-06-06 10:16:50 +0000 UTC
Labels:       <none>
Annotations:  <none>
Spec:
  PodSelector:     run=prod-nginx
  Allowing ingress traffic:
    To Port: <any> (traffic allowed to all ports)
    From: <any> (traffic not restricted by source)
  Allowing egress traffic:
    <none> (Selected pods are isolated for egress connectivity)
  Policy Types: Ingress, Egress

3. Testing egress traffic to nginx server (This is unexpected to my understanding)

Note: 10.39.0.5 is the IP of the nginx server running in 'test' namespace

Command : kubectl -n prod exec -it prod-nginx -- curl http://10.39.0.5 | grep successfully #egress
Response: <p>If you see this page, the nginx web server is successfully installed and

4. Restarted the weave-net pods

5. Retesting egress connection to same nginx server (expected)

Note: 10.39.0.5 is the IP of the nginx server running in 'test' namespace

Command: kubectl -n prod exec -it prod-nginx -- curl http://10.39.0.5 | grep successfully #egress**
Response: No connection

I would like to show you that restarting the weave-net Pods isn't required for NetworkPolicy to take effect.

Your test-network-policy NetworkPolicy is applied to Pods with the label run=prod-nginx in the prod Namespace and allows all ingress traffic and denies all egress traffic.

I will create an example to illustrate how it works.

First, I created the prod-nginx & prod-test Pods and tested the connectivity with no NetworkPolicy deployed:

# kubectl run prod-nginx --image=nginx -n prod
pod/prod-nginx created

# kubectl run prod-test --image=nginx -n prod
pod/prod-test created
    
# kubectl get pod -o wide -n prod
NAME         READY   STATUS    RESTARTS   AGE   IP          LABELS
prod-nginx   1/1     Running   0          37s   10.44.0.1   run=prod-nginx
prod-test    1/1     Running   0          11s   10.44.0.2   run=prod-test

# kubectl exec -it prod-nginx -n prod -- curl 10.44.0.1 | grep -i success
<p>If you see this page, the nginx web server is successfully installed and

# kubectl exec -it prod-nginx -n prod -- curl 10.44.0.2 | grep -i success
<p>If you see this page, the nginx web server is successfully installed and

Everything works fine, so let's deploy a test-network-policy NetworkPolicy and test again:

# cat netpol.yml 
apiVersion: networking.k8s.io/v1
kind: NetworkPolicy
metadata:
  name: test-network-policy
  namespace: prod
spec:
  ingress:
  - {}
  podSelector:
    matchLabels:
      run: prod-nginx
  policyTypes:
  - Ingress
  - Egress
  
# kubectl apply -f netpol.yml 
networkpolicy.networking.k8s.io/test-network-policy created

We can see that the prod-nginx Pod can't connect to other Pods but can connect to itself:
NOTE: A pod cannot block access to itself (see: Network Policies documentation)

# kubectl exec -it prod-nginx -n prod -- curl 10.44.0.1 | grep -i success
<p>If you see this page, the nginx web server is successfully installed and
   
# kubectl exec -it prod-nginx -n prod -- curl 10.44.0.2 | grep -i success
command terminated with exit code 7

Now let's create a stage-nginx Pod in the stage namespace and check if the prod-nginx Pod can connect to it:

# kubectl run stage-nginx --image=nginx -n stage
pod/stage-nginx created

# kubectl get pod -o wide -n stage
NAME          READY   STATUS    RESTARTS   AGE   IP          
stage-nginx   1/1     Running   0          20s   10.44.0.6   

# kubectl exec -it prod-nginx -n prod -- curl 10.44.0.6 | grep -i success
command terminated with exit code 7

We have verified that the egress rule is working properly and restarting the weave-net Pods is not required.