容器编排系统K8s之NetworkPolicy资源
我们知道在k8s上我们可以用名称空间来隔离多个资源，在不同名称空间下我们可以创建相同名称和相同类型的资源；有些资源还必须依赖名称空间才可以被创建；但是在k8s上，名称空间不可以隔离网络，所谓隔离网络是指不管pod创建在那个名称空间下，都可以被其他名称空间下pod访问到；如果我们想要限制对应名称空间下的pod被别的名称空间或外部客户端访问，我们可以在对应名称空间下创建NetworkPolicy资源；该资源可以限制哪些pod可以被外部客户端访问，或者哪些pod不能够被哪个或哪些名称空间的pod访问；除了能够限制对应名称空间下的pod网络入站流量，它还可以限制对应名称空间下的pod出站的流量，允许或拒绝某个或某些个pod访问别的名称空间下pod或外部某服务等；我们可以简单的理解NetworkPolicy就是对应名称空间的一道防火墙；我们把NetworkPolicy资源创建在对应名称空间下，那么它就能实现对该名称空间下的pod进行流量管控；
前文我们了解了k8s的网络插件flannel的基础工作逻辑，回顾请参考：https://www.cnblogs.com/qiuhom-1874/p/14225657.html；今天我们来聊一下k8s上的NetworkPolicy相关话题；
NetworkPolicy资源是做什么用的？
我们知道在k8s上我们可以用名称空间来隔离多个资源，在不同名称空间下我们可以创建相同名称和相同类型的资源；有些资源还必须依赖名称空间才可以被创建；但是在k8s上，名称空间不可以隔离网络，所谓隔离网络是指不管pod创建在那个名称空间下，都可以被其他名称空间下pod访问到；如果我们想要限制对应名称空间下的pod被别的名称空间或外部客户端访问，我们可以在对应名称空间下创建NetworkPolicy资源；该资源可以限制哪些pod可以被外部客户端访问，或者哪些pod不能够被哪个或哪些名称空间的pod访问；除了能够限制对应名称空间下的pod网络入站流量，它还可以限制对应名称空间下的pod出站的流量，允许或拒绝某个或某些个pod访问别的名称空间下pod或外部某服务等；我们可以简单的理解NetworkPolicy就是对应名称空间的一道防火墙；我们把NetworkPolicy资源创建在对应名称空间下，那么它就能实现对该名称空间下的pod进行流量管控；
NetworkPolicy基础工作逻辑
提示：首先networkpolicy是作用于一个名称空间下的pod资源之上而进行的流量管控；它可以通过pod选择器来选择对应名称空间下pod，满足对应选择器的条件的pod，将受到networkpolicy中定义的策略管控，默认pod选择器没有明确定义表示匹配对应名称空间下的所有pod；Egress是用来定义pod出站策略，默认显式定义的策略都是被允许出站的，没有定义表示拒绝；对于ingress也是同样的逻辑，只有显式定义的规则才被允许，没有定义就表示拒绝；对于出站流量来说，默认我们定义或不定义它都不会生效，都是允许放行的，只有显式配置对应出站规则生效，它才会生效；
示例：定义dev名称空间下的pod不允许任何客户端访问
[root@master01 ~]# cat denyall-ingress-dev.yaml
apiVersion: networking.k8s.io/v1
kind: NetworkPolicy
metadata:
name: deny-all-ingress
namespace: dev
spec:
podSelector: {}
policyTypes:
- Ingress
[root@master01 ~]#
提示：以上资源清单表示对dev名称空间下的pod的入站流量进行管控，拒绝所有客户端访问对应dev名称空间下的pod；这其中包含dev名称空间下的pod；
应用资源清单
在应用之前请确保你所部署的网络插件支持网络策略，默认flannel网络插件不支持网络策略，我们需要部署canal这个网络插件，该插件能够结合flannel插件，让其拥有calico的网络策略功能；
部署canal
[root@master01 ~]# kubectl apply -f https://docs.projectcalico.org/manifests/canal.yaml configmap/canal-config created customresourcedefinition.apiextensions.k8s.io/bgpconfigurations.crd.projectcalico.org created customresourcedefinition.apiextensions.k8s.io/bgppeers.crd.projectcalico.org created customresourcedefinition.apiextensions.k8s.io/blockaffinities.crd.projectcalico.org created customresourcedefinition.apiextensions.k8s.io/clusterinformations.crd.projectcalico.org created customresourcedefinition.apiextensions.k8s.io/felixconfigurations.crd.projectcalico.org created customresourcedefinition.apiextensions.k8s.io/globalnetworkpolicies.crd.projectcalico.org created customresourcedefinition.apiextensions.k8s.io/globalnetworksets.crd.projectcalico.org created customresourcedefinition.apiextensions.k8s.io/hostendpoints.crd.projectcalico.org created customresourcedefinition.apiextensions.k8s.io/ipamblocks.crd.projectcalico.org created customresourcedefinition.apiextensions.k8s.io/ipamconfigs.crd.projectcalico.org created customresourcedefinition.apiextensions.k8s.io/ipamhandles.crd.projectcalico.org created customresourcedefinition.apiextensions.k8s.io/ippools.crd.projectcalico.org created customresourcedefinition.apiextensions.k8s.io/kubecontrollersconfigurations.crd.projectcalico.org created customresourcedefinition.apiextensions.k8s.io/networkpolicies.crd.projectcalico.org created customresourcedefinition.apiextensions.k8s.io/networksets.crd.projectcalico.org created clusterrole.rbac.authorization.k8s.io/calico-kube-controllers created clusterrolebinding.rbac.authorization.k8s.io/calico-kube-controllers created clusterrole.rbac.authorization.k8s.io/calico-node created clusterrole.rbac.authorization.k8s.io/flannel configured clusterrolebinding.rbac.authorization.k8s.io/canal-flannel created clusterrolebinding.rbac.authorization.k8s.io/canal-calico created daemonset.apps/canal created serviceaccount/canal created deployment.apps/calico-kube-controllers created serviceaccount/calico-kube-controllers created poddisruptionbudget.policy/calico-kube-controllers created [root@master01 ~]#
提示：等待上述资源都正常运行以后，我们就可以应用资源清单，让其网络策略生效；
创建dev名称空间，并在其下创建一个deploy控制器，控制器创建3个nginx pod
[root@master01 manifests]# cat pod-demo.yaml
apiVersion: v1
kind: Namespace
metadata:
name: dev
labels:
name: dev
---
apiVersion: apps/v1
kind: Deployment
metadata:
name: nginx-dep
namespace: dev
spec:
replicas: 3
selector:
matchLabels:
app: nginx
rel: stable
template:
metadata:
labels:
app: nginx
rel: stable
spec:
containers:
- name: nginx
image: nginx:1.14-alpine
imagePullPolicy: IfNotPresent
ports:
- name: http
containerPort: 80
[root@master01 manifests]# kubectl apply -f pod-demo.yaml
namespace/dev unchanged
deployment.apps/nginx-dep created
[root@master01 manifests]# kubectl get pods -n dev -o wide
NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES
nginx-dep-cb74c595f-5d9tb 1/1 Running 0 51s 10.244.4.6 node04.k8s.org <none> <none>
nginx-dep-cb74c595f-ftmn9 1/1 Running 0 51s 10.244.1.5 node01.k8s.org <none> <none>
nginx-dep-cb74c595f-jf9p9 1/1 Running 0 51s 10.244.3.4 node03.k8s.org <none> <none>
[root@master01 manifests]#
在没有应用上述网络策略前，用default里的pod访问对应dev名称空间下pod，看看是否可以正常访问？
[root@master01 manifests]# kubectl get pods -n dev -o wide
NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES
nginx-dep-cb74c595f-5d9tb 1/1 Running 0 86s 10.244.4.6 node04.k8s.org <none> <none>
nginx-dep-cb74c595f-ftmn9 1/1 Running 0 86s 10.244.1.5 node01.k8s.org <none> <none>
nginx-dep-cb74c595f-jf9p9 1/1 Running 0 86s 10.244.3.4 node03.k8s.org <none> <none>
[root@master01 manifests]# kubectl get pods
NAME READY STATUS RESTARTS AGE
web-0 1/1 Running 6 3d16h
web-1 1/1 Running 5 3d16h
web-2 1/1 Running 6 3d16h
[root@master01 manifests]# kubectl exec web-0 -it -- /bin/sh
/ # wget -O - -q 10.244.4.6
<!DOCTYPE html>
<html>
<head>
<title>Welcome to nginx!</title>
<style>
body {
width: 35em;
margin: 0 auto;
font-family: Tahoma, Verdana, Arial, sans-serif;
}
</style>
</head>
<body>
<h1>Welcome to nginx!</h1>
<p>If you see this page, the nginx web server is successfully installed and
working. Further configuration is required.</p>
<p>For online documentation and support please refer to
<a href="http://nginx.org/">nginx.org</a>.<br/>
Commercial support is available at
<a href="http://nginx.com/">nginx.com</a>.</p>
<p><em>Thank you for using nginx.</em></p>
</body>
</html>
/ #
提示：可以看到用default名称空间下的pod，是可以正常访问dev名称空间下的pod；
应用资源清单
[root@master01 ~]# cat denyall-ingress-dev.yaml
apiVersion: networking.k8s.io/v1
kind: NetworkPolicy
metadata:
name: deny-all-ingress
namespace: dev
spec:
podSelector: {}
policyTypes:
- Ingress
[root@master01 ~]# kubectl apply -f denyall-ingress-dev.yaml
networkpolicy.networking.k8s.io/deny-all-ingress created
[root@master01 ~]# kubectl get netpol -n dev
NAME POD-SELECTOR AGE
deny-all-ingress <none> 10s
[root@master01 ~]# kubectl describe netpol deny-all-ingress -n dev
Name: deny-all-ingress
Namespace: dev
Created on: 2021-01-04 15:34:40 +0800 CST
Labels: <none>
Annotations: <none>
Spec:
PodSelector: <none> (Allowing the specific traffic to all pods in this namespace)
Allowing ingress traffic:
<none> (Selected pods are isolated for ingress connectivity)
Not affecting egress traffic
Policy Types: Ingress
[root@master01 ~]#
提示：从上面的netpol资源的详细信息中可以了解到，对应netpol资源pod选择器为none表示选择对应名称空间下的所有pod，allowing ingress traffic 也是为none表示没有任何客户端允许访问该名称空间下的pod；没有出站策略，表示拒绝所有出站流量；默认只是入站规则生效，所以这里的出站规则不生效，即不限制出站流量；policyTypes表示对应的那个策略生效，默认不指定是Ingress；
验证：使用default名称空间下的pod访问dev名称空间下的pod，看看是否还可以正常访问？
提示：可以看到现在default名称空间下的pod是不能访问dev名称空间下的pod；
验证：dev名称空间下的pod访问同名称空间下的pod，看看是否可以允许访问?
提示：可以看到在dev名称空间下的pod也没法访问dev名称空间下的pod，但是自己可以访问自己，其原因是自己访问自己，流量不会出站，在lo接口就被处理了，所以自己访问自己不受限制；
验证：使用dev名称空间下的pod访问default名称空间下的pod，看看是否被允许 ？
提示：可以看到使用dev名称空间下的pod访问default名称空间下的pod上没有问题，说明刚才的那条规则只是限制了dev名称空间下的pod的入站流量，并没有限制其出站流量；上面访问10.244.4.5和3.3响应404，是因为对应pod上的web服务没有主页；
示例：限制dev名称空间下的pod的出站流量，拒绝所有pod流量出站
[root@master01 ~]# cat denyall-Egress.yaml
apiVersion: networking.k8s.io/v1
kind: NetworkPolicy
metadata:
name: deny-all-egress
namespace: dev
spec:
podSelector: {}
policyTypes:
- Egress
[root@master01 ~]#
提示：policyTypes字段用来控制那个类型的规则生效，该字段是一个列表，默认是Ingress，表示只对入站流量做限制，如果明确手动定义为Egress表示，只限制出站流量；出站策略和入站策略的逻辑都是一样的，只要没有明且定义的，都是表示拒绝操作，只有明确定义的才表示对应允许操作；上述没有定义任何出站规则，表示拒绝所有pod访问外部服务；
应用资源清单
[root@master01 ~]# kubectl apply -f denyall-Egress.yaml
networkpolicy.networking.k8s.io/deny-all-egress created
[root@master01 ~]# kubectl get netpol -n dev
NAME POD-SELECTOR AGE
deny-all-egress <none> 10s
deny-all-ingress <none> 24m
[root@master01 ~]# kubectl describe netpol deny-all-egress -n dev
Name: deny-all-egress
Namespace: dev
Created on: 2021-01-04 15:59:16 +0800 CST
Labels: <none>
Annotations: <none>
Spec:
PodSelector: <none> (Allowing the specific traffic to all pods in this namespace)
Not affecting ingress traffic
Allowing egress traffic:
<none> (Selected pods are isolated for egress connectivity)
Policy Types: Egress
[root@master01 ~]#
验证：进入dev名称空间下的pod交互式接口，看看是否可以访问其他名称空间下的pod？
提示：可以看到现在dev名称空间下的pod访问default名称空间下的pod就不能访问了；
验证：删除入站限制，使用dev名称空间下的pod互相访问，看看是否可以正常访问呢？
[root@master01 ~]# kubectl get netpol -n dev
NAME POD-SELECTOR AGE
deny-all-egress <none> 5m50s
deny-all-ingress <none> 30m
[root@master01 ~]# kubectl delete netpol deny-all-ingress -n dev
networkpolicy.networking.k8s.io "deny-all-ingress" deleted
[root@master01 ~]# kubectl get netpol -n dev
NAME POD-SELECTOR AGE
deny-all-egress <none> 6m9s
[root@master01 ~]# kubectl get pods -n dev -o wide
NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES
nginx-dep-cb74c595f-5d9tb 1/1 Running 0 34m 10.244.4.6 node04.k8s.org <none> <none>
nginx-dep-cb74c595f-ftmn9 1/1 Running 0 34m 10.244.1.5 node01.k8s.org <none> <none>
nginx-dep-cb74c595f-jf9p9 1/1 Running 0 34m 10.244.3.4 node03.k8s.org <none> <none>
[root@master01 ~]# kubectl exec -it nginx-dep-cb74c595f-5d9tb -n dev -- /bin/sh
/ # wget -O - -q 10.244.1.5
^C
/ # wget -O - -q 10.244.3.4
^C
/ # wget -O - -q 10.244.4.6
<!DOCTYPE html>
<html>
<head>
<title>Welcome to nginx!</title>
<style>
body {
width: 35em;
margin: 0 auto;
font-family: Tahoma, Verdana, Arial, sans-serif;
}
</style>
</head>
<body>
<h1>Welcome to nginx!</h1>
<p>If you see this page, the nginx web server is successfully installed and
working. Further configuration is required.</p>
<p>For online documentation and support please refer to
<a href="http://nginx.org/">nginx.org</a>.<br/>
Commercial support is available at
<a href="http://nginx.com/">nginx.com</a>.</p>
<p><em>Thank you for using nginx.</em></p>
</body>
</html>
/ #
提示：可以看到限制了出站流量以后，在同一名称空间下的pod也是无法正常访问的；但是自己可以访问自己；
示例：限制dev名称空间下的pod的出站流量，允许dev名称空间下的pod互相访问
[root@master01 ~]# cat allow-dev-Egress.yaml
apiVersion: networking.k8s.io/v1
kind: NetworkPolicy
metadata:
name: allow-dev-egress
namespace: dev
spec:
podSelector: {}
egress:
- to:
- podSelector:
matchLabels:
app: nginx
rel: stable
ports:
- port: 80
protocol: TCP
policyTypes:
- Egress
[root@master01 ~]#
提示：限制出站流量，可以在spec.egress中来定义允许的出站的规则；其中to字段是用来描述被访问端，其值类型是一个列表对象；podSelector字段是用来描述被访问端pod选择器，只有满足对应pod选择器的条件的pod都被允许出站；spec.egress.ports字段是用来描述被访问端点端口信息；该字段为一个列表对象；这里需要注意，默认没有在规则指定其名称空间表示匹配当前netpol所在名称空间；以上清单表示dev下的pod允许访问dev名称空间下标签为app=nginx,rel=stable的pod的80端口；
删除deny-all-egress出站策略，应用其资源清单
[root@master01 ~]# kubectl get netpol -n dev
NAME POD-SELECTOR AGE
deny-all-egress <none> 26m
[root@master01 ~]# kubectl delete netpol deny-all-egress -n dev
networkpolicy.networking.k8s.io "deny-all-egress" deleted
[root@master01 ~]# kubectl apply -f allow-dev-Egress.yaml
networkpolicy.networking.k8s.io/allow-dev-egress created
[root@master01 ~]# kubectl get netpol -n dev
NAME POD-SELECTOR AGE
allow-dev-egress <none> 10s
[root@master01 ~]# kubectl describe netpol allow-dev-egress -n dev
Name: allow-dev-egress
Namespace: dev
Created on: 2021-01-04 16:34:23 +0800 CST
Labels: <none>
Annotations: <none>
Spec:
PodSelector: <none> (Allowing the specific traffic to all pods in this namespace)
Not affecting ingress traffic
Allowing egress traffic:
To Port: 80/TCP
To:
PodSelector: app=nginx,rel=stable
Policy Types: Egress
[root@master01 ~]#
验证：进入dev名称空间下pod的里，看看是否可以正常访问dev名称空间下的pod？
[root@master01 ~]# kubectl get pods -n dev -o wide
NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES
nginx-dep-cb74c595f-5d9tb 1/1 Running 0 65m 10.244.4.6 node04.k8s.org <none> <none>
nginx-dep-cb74c595f-ftmn9 1/1 Running 0 65m 10.244.1.5 node01.k8s.org <none> <none>
nginx-dep-cb74c595f-jf9p9 1/1 Running 0 65m 10.244.3.4 node03.k8s.org <none> <none>
[root@master01 ~]# kubectl exec -it -n dev nginx-dep-cb74c595f-5d9tb -- /bin/sh
/ # wget -O - -q 10.244.1.5
<!DOCTYPE html>
<html>
<head>
<title>Welcome to nginx!</title>
<style>
body {
width: 35em;
margin: 0 auto;
font-family: Tahoma, Verdana, Arial, sans-serif;
}
</style>
</head>
<body>
<h1>Welcome to nginx!</h1>
<p>If you see this page, the nginx web server is successfully installed and
working. Further configuration is required.</p>
<p>For online documentation and support please refer to
<a href="http://nginx.org/">nginx.org</a>.<br/>
Commercial support is available at
<a href="http://nginx.com/">nginx.com</a>.</p>
<p><em>Thank you for using nginx.</em></p>
</body>
</html>
/ # wget -O - -q 10.244.3.4
<!DOCTYPE html>
<html>
<head>
<title>Welcome to nginx!</title>
<style>
body {
width: 35em;
margin: 0 auto;
font-family: Tahoma, Verdana, Arial, sans-serif;
}
</style>
</head>
<body>
<h1>Welcome to nginx!</h1>
<p>If you see this page, the nginx web server is successfully installed and
working. Further configuration is required.</p>
<p>For online documentation and support please refer to
<a href="http://nginx.org/">nginx.org</a>.<br/>
Commercial support is available at
<a href="http://nginx.com/">nginx.com</a>.</p>
<p><em>Thank you for using nginx.</em></p>
</body>
</html>
/ #
提示：可以看到dev名称空间pod访问dev名称空间下的pod可以正常访问；
验证：使用dev名称空间pod访问default名称空间下，标签为app=nginx，rel=stable的pod，看看是否可以正常访问？
[root@master01 ~]# kubectl get pods -n dev NAME READY STATUS RESTARTS AGE nginx-dep-cb74c595f-27qjp 1/1 Running 2 88m nginx-dep-cb74c595f-b92s4 1/1 Running 1 88m nginx-dep-cb74c595f-wdqnh 1/1 Running 1 88m [root@master01 ~]# kubectl get pods -o wide --show-labels NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES LABELS nginx-dep-cb74c595f-5mwl2 1/1 Running 1 89m 10.244.4.18 node04.k8s.org <none> <none> app=nginx,pod-template-hash=cb74c595f,rel=stable nginx-dep-cb74c595f-vxc9z 1/1 Running 1 89m 10.244.3.12 node03.k8s.org <none> <none> app=nginx,pod-template-hash=cb74c595f,rel=stable nginx-dep-cb74c595f-z265t 1/1 Running 1 89m 10.244.2.13 node02.k8s.org <none> <none> app=nginx,pod-template-hash=cb74c595f,rel=stable web-0 1/1 Running 8 3d20h 10.244.2.14 node02.k8s.org <none> <none> app=nginx,controller-revision-hash=web-6db9455fdf,statefulset.kubernetes.io/pod-name=web-0 web-1 1/1 Running 7 3d20h 10.244.4.16 node04.k8s.org <none> <none> app=nginx,controller-revision-hash=web-6db9455fdf,statefulset.kubernetes.io/pod-name=web-1 web-2 1/1 Running 8 3d20h 10.244.3.11 node03.k8s.org <none> <none> app=nginx,controller-revision-hash=web-6db9455fdf,statefulset.kubernetes.io/pod-name=web-2 [root@master01 ~]# kubectl exec -it nginx-dep-cb74c595f-27qjp -n dev -- /bin/sh / # wget -O - -q 10.244.4.18 ^C / # wget -O - -q 10.244.3.12 ^C / # wget -O - -q 10.244.2.13 ^C / #
提示：可以看到dev名称空间下的pod访问default名称空间下标签为app=nginx，rel=stable的pod也是无法正常访问，原因是我们在策略中没有指定对应访问目标的名称空间，默认就是对应netpol所在名称空间；
示例：允许dev名称空间下的pod访问default和dev名称空间下标签为app=nginx，rel=stable的pod
[root@master01 ~]# cat allow-dev-def-egress.yaml
apiVersion: networking.k8s.io/v1
kind: NetworkPolicy
metadata:
name: allow-dev-egress
namespace: dev
spec:
podSelector: {}
egress:
- to:
- namespaceSelector:
matchExpressions:
- key: name
operator: In
values: ["def","dev"]
podSelector:
matchLabels:
app: nginx
rel: stable
ports:
- port: 80
protocol: TCP
policyTypes:
- Egress
[root@master01 ~]#
提示：spec.egress.to中的namespeceSelector字段使用来描述被访问端所在名称空间；podSelector字段是用来描述被访问端的pod；如果这两个字段中只有一个字段前用了“-”，那么namespaceSelector和podSelector所指定的条件是与关系，即两个条件必须都满足；如果两个字段都用了“-”开头，则两个字段所指定的条件就是或关系，即满足其中一个字段的条件都可以被规则匹配；上述示例表示dev下的所有pod能够访问名称空间上的标签是name=def或name=dev的名称空间下的pod，并且这些pod的标签必须是app=nginx,rel=stable；即要想被dev名称空间下的pod所访问，对应pod的应该同时满足两个条件，第一对应名称空间上有name=def或者name=dev的标签；其次pod本身要有两个标签，app=nginx和rel=stable；这两个条件缺一不可；
删除原有netpol，应用新的清单
[root@master01 ~]# kubectl get netpol -n dev
NAME POD-SELECTOR AGE
allow-dev-egress <none> 28m
[root@master01 ~]# kubectl delete netpol allow-dev-egress -n dev
networkpolicy.networking.k8s.io "allow-dev-egress" deleted
[root@master01 ~]# kubectl apply -f allow-dev-def-egress.yaml
networkpolicy.networking.k8s.io/allow-dev-egress created
[root@master01 ~]# kubectl get netpol -n dev
NAME POD-SELECTOR AGE
allow-dev-egress <none> 5s
[root@master01 ~]# kubectl describe netpol allow-dev-egress -n dev
Name: allow-dev-egress
Namespace: dev
Created on: 2021-01-04 19:44:32 +0800 CST
Labels: <none>
Annotations: <none>
Spec:
PodSelector: <none> (Allowing the specific traffic to all pods in this namespace)
Not affecting ingress traffic
Allowing egress traffic:
To Port: 80/TCP
To:
NamespaceSelector: name in (def,dev)
PodSelector: app=nginx,rel=stable
Policy Types: Egress
[root@master01 ~]#
提示：可以看到对应的出站规则就有两个条件，首先称空间必须匹配，name=def或者name=dev，即对应名称空间必须要这样的标签；其次pod必须满足有app=nginx,rel=stable的标签；
给default名称空间打上name=def的标签
[root@master01 ~]# kubectl get ns --show-labels NAME STATUS AGE LABELS default Active 27d <none> dev Active 4h24m name=dev ingress-nginx Active 13d app.kubernetes.io/name=ingress-nginx,app.kubernetes.io/part-of=ingress-nginx kube-node-lease Active 27d <none> kube-public Active 27d <none> kube-system Active 27d <none> kubernetes-dashboard Active 2d5h <none> prod Active 57m name=prod [root@master01 ~]# kubectl label ns default name=def namespace/default labeled [root@master01 ~]# kubectl get ns --show-labels NAME STATUS AGE LABELS default Active 27d name=def dev Active 4h24m name=dev ingress-nginx Active 13d app.kubernetes.io/name=ingress-nginx,app.kubernetes.io/part-of=ingress-nginx kube-node-lease Active 27d <none> kube-public Active 27d <none> kube-system Active 27d <none> kubernetes-dashboard Active 2d5h <none> prod Active 58m name=prod [root@master01 ~]#
验证：用dev下的pod访问default和dev名称空间下pod标签为app=nginx,rel=stable的pod，看看对应pod是否能够被访问到？
[root@master01 ~]# kubectl get pods -n dev -o wide --show-labels NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES LABELS nginx-dep-cb74c595f-27qjp 1/1 Running 2 116m 10.244.1.18 node01.k8s.org <none> <none> app=nginx,pod-template-hash=cb74c595f,rel=stable nginx-dep-cb74c595f-b92s4 1/1 Running 1 116m 10.244.4.17 node04.k8s.org <none> <none> app=nginx,pod-template-hash=cb74c595f,rel=stable nginx-dep-cb74c595f-wdqnh 1/1 Running 1 116m 10.244.1.19 node01.k8s.org <none> <none> app=nginx,pod-template-hash=cb74c595f,rel=stable [root@master01 ~]# kubectl get pods -o wide --show-labels NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES LABELS nginx-dep-cb74c595f-5mwl2 1/1 Running 1 117m 10.244.4.18 node04.k8s.org <none> <none> app=nginx,pod-template-hash=cb74c595f,rel=stable nginx-dep-cb74c595f-vxc9z 1/1 Running 1 117m 10.244.3.12 node03.k8s.org <none> <none> app=nginx,pod-template-hash=cb74c595f,rel=stable nginx-dep-cb74c595f-z265t 1/1 Running 1 117m 10.244.2.13 node02.k8s.org <none> <none> app=nginx,pod-template-hash=cb74c595f,rel=stable web-0 1/1 Running 8 3d21h 10.244.2.14 node02.k8s.org <none> <none> app=nginx,controller-revision-hash=web-6db9455fdf,statefulset.kubernetes.io/pod-name=web-0 web-1 1/1 Running 7 3d21h 10.244.4.16 node04.k8s.org <none> <none> app=nginx,controller-revision-hash=web-6db9455fdf,statefulset.kubernetes.io/pod-name=web-1 web-2 1/1 Running 8 3d21h 10.244.3.11 node03.k8s.org <none> <none> app=nginx,controller-revision-hash=web-6db9455fdf,statefulset.kubernetes.io/pod-name=web-2 [root@master01 ~]# kubectl exec -it -n dev nginx-dep-cb74c595f-27qjp -- /bin/sh / # wget --spider --timeout=1 10.244.4.17 Connecting to 10.244.4.17 (10.244.4.17:80) / # wget --spider --timeout=1 10.244.1.19 Connecting to 10.244.1.19 (10.244.1.19:80) / # wget --spider --timeout=1 10.244.4.18 Connecting to 10.244.4.18 (10.244.4.18:80) / # wget --spider --timeout=1 10.244.3.12 Connecting to 10.244.3.12 (10.244.3.12:80) / # wget --spider --timeout=1 10.244.2.13 Connecting to 10.244.2.13 (10.244.2.13:80) / # wget --spider --timeout=1 10.244.2.14 Connecting to 10.244.2.14 (10.244.2.14:80) wget: download timed out / #
提示：可以看到用dev名称空间的pod访问名称空间上有name=def或name=dev标签并且对应pod上有app=nginx和rel=stable的标签的pod是能够正常访问到，如果不能满足对应名称空间和pod标签的pod就不能被访问；上面示例中最后一个podip不满足pod的标签，所以访问不到对应pod；
示例：允许dev名称空间下的pod访问名称空间上有name=def或name=dev标签的名称空间下的pod或者名称空间上有name=prod标签，并且对应名称空间下的pod必须有app=nginx，rel=stable的标签；
[root@master01 ~]# cat allow-dev-def-egress.yaml
apiVersion: networking.k8s.io/v1
kind: NetworkPolicy
metadata:
name: allow-dev-egress
namespace: dev
spec:
podSelector: {}
egress:
- to:
- namespaceSelector:
matchExpressions:
- key: name
operator: In
values: ["def","dev"]
- to:
- namespaceSelector:
matchLables:
name: prod
podSelector:
matchLabels:
app: nginx
rel: stable
ports:
- port: 80
protocol: TCP
policyTypes:
- Egress
[root@master01 ~]#
应用配置清单
[root@master01 ~]# kubectl apply -f allow-dev-def-egress.yaml
networkpolicy.networking.k8s.io/allow-dev-egress created
[root@master01 ~]# kubectl get netpol -n dev
NAME POD-SELECTOR AGE
allow-dev-egress <none> 10s
[root@master01 ~]# kubectl describe netpol allow-dev-egress -n dev
Name: allow-dev-egress
Namespace: dev
Created on: 2021-01-04 20:44:42 +0800 CST
Labels: <none>
Annotations: <none>
Spec:
PodSelector: <none> (Allowing the specific traffic to all pods in this namespace)
Not affecting ingress traffic
Allowing egress traffic:
To Port: <any> (traffic allowed to all ports)
To:
NamespaceSelector: name in (def,dev)
----------
To Port: 80/TCP
To:
NamespaceSelector: name=prod
PodSelector: app=nginx,rel=stable
Policy Types: Egress
[root@master01 ~]#
验证：用dev名称空间下的pod访问dev名称空间下的pod，看看是否可以正常访问？
[root@master01 ~]# kubectl get pods -n dev -o wide NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES nginx-dep-cb74c595f-27qjp 1/1 Running 2 168m 10.244.1.18 node01.k8s.org <none> <none> nginx-dep-cb74c595f-b92s4 1/1 Running 1 168m 10.244.4.17 node04.k8s.org <none> <none> nginx-dep-cb74c595f-wdqnh 1/1 Running 1 168m 10.244.1.19 node01.k8s.org <none> <none> [root@master01 ~]# kubectl exec -it -n dev nginx-dep-cb74c595f-27qjp -- /bin/sh / # wget --spider --timeout=1 10.244.4.17 Connecting to 10.244.4.17 (10.244.4.17:80) / # wget --spider --timeout=1 10.244.1.19 Connecting to 10.244.1.19 (10.244.1.19:80) / # exit [root@master01 ~]#
提示：可以看到dev名称空间下的pod相互访问没有问题；
验证：用dev名称空间下的pod访问default名称空间下的pod ，看看是否可以正常访问？
[root@master01 ~]# kubectl get pods -n dev -o wide NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES nginx-dep-cb74c595f-27qjp 1/1 Running 2 170m 10.244.1.18 node01.k8s.org <none> <none> nginx-dep-cb74c595f-b92s4 1/1 Running 1 170m 10.244.4.17 node04.k8s.org <none> <none> nginx-dep-cb74c595f-wdqnh 1/1 Running 1 170m 10.244.1.19 node01.k8s.org <none> <none> [root@master01 ~]# kubectl get pods -o wide NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES nginx-dep-cb74c595f-5mwl2 1/1 Running 1 171m 10.244.4.18 node04.k8s.org <none> <none> nginx-dep-cb74c595f-vxc9z 1/1 Running 1 171m 10.244.3.12 node03.k8s.org <none> <none> nginx-dep-cb74c595f-z265t 1/1 Running 1 171m 10.244.2.13 node02.k8s.org <none> <none> web-0 1/1 Running 8 3d22h 10.244.2.14 node02.k8s.org <none> <none> web-1 1/1 Running 7 3d22h 10.244.4.16 node04.k8s.org <none> <none> web-2 1/1 Running 8 3d22h 10.244.3.11 node03.k8s.org <none> <none> [root@master01 ~]# kubectl exec -it -n dev nginx-dep-cb74c595f-27qjp -- /bin/sh / # wget --spider --timeout=1 10.244.4.18 Connecting to 10.244.4.18 (10.244.4.18:80) / # wget --spider --timeout=1 10.244.3.12 Connecting to 10.244.3.12 (10.244.3.12:80) / # wget --spider --timeout=1 10.244.2.13 Connecting to 10.244.2.13 (10.244.2.13:80) / # wget --spider --timeout=1 10.244.2.14 Connecting to 10.244.2.14 (10.244.2.14:80) / # wget --spider --timeout=1 10.244.4.16 Connecting to 10.244.4.16 (10.244.4.16:80) wget: server returned error: HTTP/1.1 403 Forbidden / # wget --spider --timeout=1 10.244.3.11 Connecting to 10.244.3.11 (10.244.3.11:80) wget: server returned error: HTTP/1.1 403 Forbidden / #
提示：可以看到default名称空间下的所有pod都能够被dev名称空间下的pod所访问；后面显示403是因为对应web服务没有主页；
验证：使用dev名称空间下的pod访问prod名称空间下的pod，看看对应pod是否能够被访问？
[root@master01 ~]# kubectl get ns prod --show-labels NAME STATUS AGE LABELS prod Active 118m name=prod [root@master01 ~]# kubectl get pods -n prod -o wide --show-labels NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES LABELS nginx-dep-cb74c595f-rzbv2 1/1 Running 0 118m 10.244.4.19 node04.k8s.org <none> <none> app=nginx,pod-template-hash=cb74c595f,rel=stable nginx-dep-cb74c595f-v8ssx 1/1 Running 0 118m 10.244.3.13 node03.k8s.org <none> <none> app=nginx,pod-template-hash=cb74c595f,rel=stable nginx-dep-cb74c595f-zsqcc 1/1 Running 0 118m 10.244.2.15 node02.k8s.org <none> <none> app=nginx,pod-template-hash=cb74c595f,rel=stable nginx-dep1-f85fdcdbc-kdq5b 1/1 Running 0 117m 10.244.1.21 node01.k8s.org <none> <none> app=nginx,pod-template-hash=f85fdcdbc nginx-dep1-f85fdcdbc-n8cvs 1/1 Running 0 117m 10.244.3.14 node03.k8s.org <none> <none> app=nginx,pod-template-hash=f85fdcdbc nginx-dep1-f85fdcdbc-vz2mp 1/1 Running 0 117m 10.244.4.20 node04.k8s.org <none> <none> app=nginx,pod-template-hash=f85fdcdbc [root@master01 ~]# kubectl get pods -n dev NAME READY STATUS RESTARTS AGE nginx-dep-cb74c595f-27qjp 1/1 Running 2 174m nginx-dep-cb74c595f-b92s4 1/1 Running 1 174m nginx-dep-cb74c595f-wdqnh 1/1 Running 1 174m [root@master01 ~]# kubectl exec -it -n dev nginx-dep-cb74c595f-27qjp -- /bin/sh / # wget --spider --timeout=1 10.244.4.20 Connecting to 10.244.4.20 (10.244.4.20:80) wget: download timed out / # wget --spider --timeout=1 10.244.3.14 Connecting to 10.244.3.14 (10.244.3.14:80) wget: download timed out / # wget --spider --timeout=1 10.244.1.21 Connecting to 10.244.1.21 (10.244.1.21:80) wget: download timed out / # wget --spider --timeout=1 10.244.2.15 Connecting to 10.244.2.15 (10.244.2.15:80) / # wget --spider --timeout=1 10.244.3.13 Connecting to 10.244.3.13 (10.244.3.13:80) / # wget --spider --timeout=1 10.244.4.19 Connecting to 10.244.4.19 (10.244.4.19:80) / #
提示：可以看到在prod名称空间下对应pod标签为app=nginx,rel=stable的pod都能够被dev下的pod所访问，如果对应标签不是app=nginx,rel=stable的pod则不能被访问到；
从上面的示例可以总结，在k8s上网络策略是白名单机制，所谓白名单机制是指，只有明确定义的策略才会被允许放行，默认没有指定的规则就是拒绝的，即条件不匹配的都会被拒绝；其次对于ingress或egress来说，对应的from或to都是用来指定访问端或被访问端的信息；如果我们在对应的字段中没有定义namespaceSelector字段，默认ingress或egrss会匹配当前netpol所在名称空间，即在没有明确指定namespaceSelector字段时，对应的其他条件都是针对当前netpol所在名称空间；多个条件组合使用，如果多个条件都在一个列表中，则表示多个条件间是与关系，即指定的条件需要同时满足对应策略才会放行；如果多个条件不再同一个列表中，则多个条件之间是或关系，即满足其中一个条件都会被对应策略放行；
作者：Linux-1874
本文版权归作者和博客园共有，欢迎转载，但未经作者同意必须保留此段声明，且在文章页面明显位置给出原文连接，否则保留追究法律责任的权利.
