十七、Kubernetes之Calico网络策略
一、网络策略
网络策略(Network Policy )是 Kubernetes 的一种资源。Network Policy 通过 Label 选择 Pod,并指定其他 Pod 或外界如何与这些 Pod 通信。
Pod的网络流量包含流入(Ingress)和流出(Egress)两种方向。默认情况下,所有 Pod 是非隔离的,即任何来源的网络流量都能够访问 Pod,没有任何限制。当为 Pod 定义了 Network Policy,只有 Policy 允许的流量才能访问 Pod。
Kubernetes的网络策略功能也是由第三方的网络插件实现的,因此,只有支持网络策略功能的网络插件才能进行配置网络策略,比如Calico、Canal、kube-router等等。
二、配置网络策略
在Kubernetes系统中,报文的流入和流出的核心组件是Pod资源,它们也是网络策略功能的主要应用对象。NetworkPolicy
对象通过podSelector
选择 一组Pod资源作为控制对象。NetworkPolicy
是定义在一组Pod资源之上用于管理入站流量,或出站流量的一组规则,有可以是出入站规则一起生效,规则的生效模式通常由spec.policyTypes
进行 定义。如下图:
默认情况下,Pod对象的流量控制是为空的,报文可以自由出入。在附加网络策略之后,Pod对象会因为NetworkPolicy而被隔离,一旦名称空间中有任何NetworkPolicy对象匹配了某特定的Pod对象,则该Pod将拒绝NetworkPolicy规则中不允许的所有连接请求,但是那些未被匹配到的Pod对象依旧可以接受所有流量。
就特定的Pod集合来说,入站和出站流量默认是放行状态,除非有规则可以进行匹配。还有一点需要注意的是,在spec.policyTypes
中指定了生效的规则类型,但是在networkpolicy.spec
字段中嵌套定义了没有任何规则的Ingress或Egress时,则表示拒绝入站或出站的一切流量。定义网络策略的基本格式如下:
apiVersion: networking.k8s.io/v1 #定义API版本 kind: NetworkPolicy #定义资源类型 metadata: name: allow-myapp-ingress #定义NetwokPolicy的名字 namespace: default spec: #NetworkPolicy规则定义 podSelector: #匹配拥有标签app:myapp的Pod资源 matchLabels: app: myapp policyTypes ["Ingress"] #NetworkPolicy类型,可以是Ingress,Egress,或者两者共存 ingress: #定义入站规则 - from: - ipBlock: #定义可以访问的网段 cidr: 10.244.0.0/16 except: #排除的网段 - 10.244.3.0/24 - podSelector: #选定当前default名称空间,标签为app:myapp可以入站 matchLabels: app: myapp ports: #开放的协议和端口定义 - protocol: TCP port: 80 该网络策略就是将default名称空间中拥有标签"app=myapp"的Pod资源开放80/TCP端口给10.244.0.0/16网段,并排除10.244.3.0/24网段的访问,并且也开放给标签为app=myapp的所有Pod资源进行访问。
三、实验案例
环境准备
创建linux和python两个ns并分别在这两个ns中创建pod资源来用于测试网络策略
root@k8s-deploy-harbor:~/manifest/network-policy# kubectl create ns linux namespace/linux created root@k8s-deploy-harbor:~/manifest/network-policy# kubectl create ns python namespace/python created root@k8s-deploy-harbor:~/manifest/network-policy
创建linux namespace中的资源
nginx-linux-ns.yaml
apiVersion: apps/v1 kind: Deployment metadata: labels: app: linux-nginx-deployment-label name: linux-nginx-deployment namespace: linux spec: replicas: 1 selector: matchLabels: app: linux-nginx-selector template: metadata: labels: app: linux-nginx-selector spec: containers: - name: linux-nginx-container image: harbor.zhrx.com/baseimages/nginx:alpine imagePullPolicy: IfNotPresent ports: - containerPort: 80 protocol: TCP name: http - containerPort: 443 protocol: TCP name: https --- kind: Service apiVersion: v1 metadata: labels: app: linux-nginx-service-label name: linux-nginx-service namespace: linux spec: type: NodePort ports: - name: http port: 80 protocol: TCP targetPort: 80 nodePort: 30004 - name: https port: 443 protocol: TCP targetPort: 443 nodePort: 30443 selector: app: linux-nginx-selector
tomcat-linux-ns.yaml
apiVersion: apps/v1 kind: Deployment metadata: labels: app: linux-tomcat-app1-deployment-label name: linux-tomcat-app1-deployment namespace: linux spec: replicas: 1 selector: matchLabels: app: linux-tomcat-app1-selector template: metadata: labels: app: linux-tomcat-app1-selector spec: containers: - name: linux-tomcat-app1-container image: harbor.zhrx.com/baseimages/tomcat:7.0.94-alpine imagePullPolicy: IfNotPresent ports: - containerPort: 8080 protocol: TCP name: http --- kind: Service apiVersion: v1 metadata: labels: app: linux-tomcat-app1-service-label name: linux-tomcat-app1-service namespace: linux spec: type: NodePort ports: - name: http port: 80 protocol: TCP targetPort: 8080 nodePort: 30005 selector: app: linux-tomcat-app1-selector
创建资源
root@k8s-deploy-harbor:~/manifest/network-policy/linux-ns# kubectl apply -f nginx-linux-ns.yaml deployment.apps/linux-nginx-deployment created service/linux-nginx-service created root@k8s-deploy-harbor:~/manifest/network-policy/linux-ns# kubectl apply -f tomcat-linux-ns.yaml deployment.apps/linux-tomcat-app1-deployment created service/linux-tomcat-app1-service created root@k8s-deploy-harbor:~/manifest/network-policy/linux-ns# kubectl get pod,svc -n linux NAME READY STATUS RESTARTS AGE pod/linux-nginx-deployment-74fc689864-f78nb 1/1 Running 0 4m3s pod/linux-tomcat-app1-deployment-7cc6bdd987-27wf5 1/1 Running 0 14s NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE service/linux-nginx-service NodePort 10.68.187.8 <none> 80:30004/TCP,443:30443/TCP 4m3s service/linux-tomcat-app1-service NodePort 10.68.170.248 <none> 80:30005/TCP 14s root@k8s-master:~# kubectl exec -it linux-tomcat-app1-deployment-7cc6bdd987-27wf5 -n linux sh kubectl exec [POD] [COMMAND] is DEPRECATED and will be removed in a future version. Use kubectl exec [POD] -- [COMMAND] instead. /usr/local/tomcat # mkdir webapps/linux /usr/local/tomcat # echo "linux tomcat" > webapps/linux/index.jsp
创建python namespace中的资源
nginx-python-ns.yaml
apiVersion: apps/v1 kind: Deployment metadata: labels: app: python-nginx-deployment-label name: python-nginx-deployment namespace: python spec: replicas: 1 selector: matchLabels: app: python-nginx-selector template: metadata: labels: app: python-nginx-selector project: python spec: containers: - name: python-nginx-container image: harbor.zhrx.com/baseimages/nginx:alpine imagePullPolicy: IfNotPresent ports: - containerPort: 80 protocol: TCP name: http - containerPort: 443 protocol: TCP name: https --- kind: Service apiVersion: v1 metadata: labels: app: python-nginx-service-label name: python-nginx-service namespace: python spec: type: NodePort ports: - name: http port: 80 protocol: TCP targetPort: 80 nodePort: 30014 - name: https port: 443 protocol: TCP targetPort: 443 nodePort: 30453 selector: app: python-nginx-selector project: python #一个或多个selector,至少能匹配目标pod的一个标签
tomcat-python-ns.yaml
apiVersion: apps/v1 kind: Deployment metadata: labels: app: python-tomcat-app1-deployment-label name: python-tomcat-app1-deployment namespace: python spec: replicas: 1 selector: matchLabels: app: python-tomcat-app1-selector template: metadata: labels: app: python-tomcat-app1-selector spec: containers: - name: python-tomcat-app1-container image: harbor.zhrx.com/baseimages/tomcat:7.0.94-alpine imagePullPolicy: IfNotPresent ports: - containerPort: 8080 protocol: TCP name: http --- kind: Service apiVersion: v1 metadata: labels: app: python-tomcat-app1-service-label name: python-tomcat-app1-service namespace: python spec: type: NodePort ports: - name: http port: 80 protocol: TCP targetPort: 8080 nodePort: 30015 selector: app: python-tomcat-app1-selector
创建资源
root@k8s-deploy-harbor:~/manifest/network-policy/python-ns# kubectl apply -f nginx-python-ns.yaml deployment.apps/python-nginx-deployment created service/python-nginx-service created root@k8s-deploy-harbor:~/manifest/network-policy/python-ns# kubectl apply -f tomcat-python-ns.yaml deployment.apps/python-tomcat-app1-deployment created service/python-tomcat-app1-service created root@k8s-deploy-harbor:~/manifest/network-policy/python-ns# kubectl get pod,svc -n python NAME READY STATUS RESTARTS AGE pod/python-nginx-deployment-889b9fd95-6c9pv 1/1 Running 0 11s pod/python-tomcat-app1-deployment-85d7bcb88c-xcrz4 1/1 Running 0 7s NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE service/python-nginx-service NodePort 10.68.192.11 <none> 80:30014/TCP,443:30453/TCP 10s service/python-tomcat-app1-service NodePort 10.68.88.145 <none> 80:30015/TCP 7s root@k8s-master:~# kubectl exec -it python-tomcat-app1-deployment-85d7bcb88c-xcrz4 -n python sh kubectl exec [POD] [COMMAND] is DEPRECATED and will be removed in a future version. Use kubectl exec [POD] -- [COMMAND] instead. /usr/local/tomcat # mkdir webapps/python /usr/local/tomcat # echo "python tomcat" > webapps/python/index.jsp
在default namespace中创建测试pod
testpod-default.yaml
apiVersion: apps/v1 kind: Deployment metadata: labels: app: testpod-busybox-deployment-label name: testpod-busybox-deployment namespace: default spec: replicas: 1 selector: matchLabels: app: python-nginx-selector project: python template: metadata: labels: app: python-nginx-selector project: python spec: containers: - name: testpod-busybox-container image: busybox:latest command: - sleep - "50000000" imagePullPolicy: IfNotPresentc
创建资源
root@k8s-deploy-harbor:~/manifest/network-policy/python-ns# kubectl apply -f testpod-busybox.yaml deployment.apps/testpod-busybox-deployment created root@k8s-deploy-harbor:~/manifest/network-policy/python-ns# kubectl get pod NAME READY STATUS RESTARTS AGE testpod-busybox-deployment-8f4f66686-wvw6c 1/1 Running 0 49s
case1-ingress-podSelector
仅允许python namespace下拥有project: "python"标签的pod访问 拥有app: python-tomcat-app1-selector 标签的pod
apiVersion: networking.k8s.io/v1 kind: NetworkPolicy metadata: name: tomcat-access--networkpolicy namespace: python spec: policyTypes: - Ingress podSelector: matchLabels: app: python-tomcat-app1-selector #对匹配到的目的Pod应用以下规则 ingress: #入栈规则,如果指定目标端口就是匹配全部端口和协议,协议TCP, UDP, or SCTP - from: - podSelector: matchLabels: #app: python-nginx-selector #如果存在多个matchLabel条件,是or的关系,即要同时满足条件A、条件B、条件X project: "python"
创建规则并验证
root@k8s-deploy-harbor:~/manifest/network-policy/python-ns# kubectl apply -f case1-ingress-podSelector.yaml networkpolicy.networking.k8s.io/tomcat-access--networkpolicy created #在python ns下的nginx pod中可以访问到tomcat root@k8s-master:~# kubectl get svc -n python NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE python-nginx-service NodePort 10.68.192.11 <none> 80:30014/TCP,443:30453/TCP 39m python-tomcat-app1-service NodePort 10.68.88.145 <none> 80:30015/TCP 39m root@k8s-master:~# kubectl exec -it python-nginx-deployment-889b9fd95-6c9pv -n python sh kubectl exec [POD] [COMMAND] is DEPRECATED and will be removed in a future version. Use kubectl exec [POD] -- [COMMAND] instead. / # curl python-tomcat-app1-service/python/ python tomcat #在linux ns下的nginx pod中不可访问 root@k8s-master:~# kubectl exec -it linux-nginx-deployment-74fc689864-f78nb -n linux sh kubectl exec [POD] [COMMAND] is DEPRECATED and will be removed in a future version. Use kubectl exec [POD] -- [COMMAND] instead. / # curl python-tomcat-app1-service.python.svc.cluster.local/python/ / # 阻塞无返回
case2-ingress-podSelector-ns-SinglePort
仅允许python namespace下拥有project: "python"标签的pod访问 拥有app: python-tomcat-app1-selector 标签的pod 的8080端口
apiVersion: networking.k8s.io/v1 kind: NetworkPolicy metadata: name: tomcat-access--networkpolicy namespace: python spec: policyTypes: - Ingress podSelector: matchLabels: app: python-tomcat-app1-selector ingress: - from: - podSelector: matchLabels: #app: python-nginx-selector #指定访问源的匹配条件,如果存在多个matchLabel条件,是or的关系,即要同时满足条件A、条件B、条件X project: "python" ports: #入栈规则,如果指定目标端口就是匹配全部端口和协议,协议TCP, UDP, or SCTP - protocol: TCP port: 8080 #允许通过TCP协议访问目标pod的8080端口,但是其它没有允许的端口将全部禁止访问 #port: 80
创建规则并验证
root@k8s-deploy-harbor:~/manifest/network-policy/python-ns# kubectl apply -f case2-ingress-podSelector-ns-SinglePort.yaml networkpolicy.networking.k8s.io/tomcat-access--networkpolicy created # 通ns下nginx pod可以访问tomcat pod的8080端口 root@k8s-master:~# kubectl exec -it python-nginx-deployment-889b9fd95-6c9pv -n python sh kubectl exec [POD] [COMMAND] is DEPRECATED and will be removed in a future version. Use kubectl exec [POD] -- [COMMAND] instead. / # curl 172.20.36.69:8080/python/ python tomcat
case3-ingress-podSelector-ns-MultiPort
仅允许python namespace下拥有project: "python"标签的pod访问 拥有app: python-tomcat-app1-selector 标签的pod 的8080和8009端口
apiVersion: networking.k8s.io/v1 kind: NetworkPolicy metadata: name: tomcat-access--networkpolicy namespace: python spec: policyTypes: - Ingress podSelector: #目标pod matchLabels: app: python-tomcat-app1-selector ingress: - from: - podSelector: #匹配源pod,matchLabels: {}为不限制源pod即允许所有pod,写法等同于resources(不加就是不限制) matchLabels: {} ports: #入栈规则,如果指定目标端口就是匹配全部端口和协议,协议TCP, UDP, or SCTP - protocol: TCP port: 8080 #允许通过TCP协议访问目标pod的8080端口,但是其它没有允许的端口将全部禁止访问 - protocol: TCP port: 8009
case4-ingress-podSelector-ns
允许当前namespace下所有pod的通信
apiVersion: networking.k8s.io/v1 kind: NetworkPolicy metadata: name: tomcat-access--networkpolicy namespace: python spec: policyTypes: - Ingress podSelector: #目标pod matchLabels: {} #匹配所有目标pod ingress: - from: - podSelector: #匹配源pod,matchLabels: {}为不限制源pod即允许所有pod,写法等同于resources(不加就是不限制) matchLabels: {}
case5-ingress-ipBlock
允许指定的ip网段访问tomcat pod的808端口
apiVersion: networking.k8s.io/v1 kind: NetworkPolicy metadata: name: tomcat-access--networkpolicy namespace: python spec: policyTypes: - Ingress podSelector: #目标pod matchLabels: app: python-tomcat-app1-selector ingress: - from: - ipBlock: cidr: 10.200.0.0/16 #白名单,允许访问的地址范围,没有允许的将禁止访问目标pod except: - 10.200.218.0/24 #在以上范围内禁止访问的源IP地址 - 10.200.229.0/24 #在以上范围内禁止访问的源IP地址 - 10.200.230.239/32 #在以上范围内禁止访问的源IP地址 ports: #入栈规则,如果指定目标端口就是匹配全部端口和协议,协议TCP, UDP, or SCTP - protocol: TCP port: 8080 #允许通过TCP协议访问目标pod的8080端口,但是其它没有允许的端口将全部禁止访问
case6-ingress-namespaceSelector
允许指定的ns访问当前ns下的所有pod 的某些端口
apiVersion: networking.k8s.io/v1 kind: NetworkPolicy metadata: name: tomcat-access--networkpolicy namespace: python spec: policyTypes: - Ingress podSelector: #目标pod matchLabels: {} #允许访问python namespace 中的所有pod # app: python-tomcat-app1-selector #可以只允许访问python namespace中指定的pod ingress: - from: - namespaceSelector: matchLabels: nsname: linux #只允许指定的namespace访问 - namespaceSelector: matchLabels: nsname: python #只允许指定的namespace访问 ports: #入栈规则,如果指定目标端口就是匹配全部端口和协议,协议TCP, UDP, or SCTP - protocol: TCP port: 8080 #允许通过TCP协议访问目标pod的8080端口,但是其它没有允许的端口将全部禁止访问 #port: 80 - protocol: TCP port: 3306 - protocol: TCP port: 6379
case7-Egress-ipBlock
允许指定pod访问目的ip网段的指定端口
apiVersion: networking.k8s.io/v1 kind: NetworkPolicy metadata: name: egress-access-networkpolicy namespace: python spec: policyTypes: - Egress podSelector: #目标pod选择器 matchLabels: #基于label匹配目标pod app: python-tomcat-app1-selector #匹配python namespace中app的值为python-tomcat-app1-selector的pod,然后基于egress中的指定网络策略进行出口方向的网络限制 egress: - to: - ipBlock: cidr: 10.200.0.0/16 #允许匹配到的pod出口访问的目的CIDR地址范围 - ipBlock: cidr: 172.31.7.106/32 #允许匹配到的pod出口访问的目的主机 ports: - protocol: TCP port: 80 #允许匹配到的pod访问目的端口为80的访问 - protocol: TCP port: 53 #允许匹配到的pod访问目的端口为53 即DNS的解析 - protocol: UDP port: 53 #允许匹配到的pod访问目的端口为53 即DNS的解析
case8-Egress-PodSelector
允许拥有app: python-nginx-selector标签的pod 访问 拥有app: python-tomcat-app1-selector 的pod的指定端口
apiVersion: networking.k8s.io/v1 kind: NetworkPolicy metadata: name: egress-access-networkpolicy namespace: python spec: policyTypes: - Egress podSelector: #目标pod选择器 matchLabels: #基于label匹配目标pod app: python-nginx-selector #匹配python namespace中app的值为python-tomcat-app1-selector的pod,然后基于egress中的指定网络策略进行出口方向的网络限制 egress: - to: - podSelector: #匹配pod,matchLabels: {}为不限制源pod即允许所有pod,写法等同于resources(不加就是不限制) matchLabels: app: python-tomcat-app1-selector ports: - protocol: TCP port: 8080 #允许80端口的访问 - protocol: TCP port: 53 #允许DNS的解析 - protocol: UDP port: 53
case9-Egress-namespaceSelector
允许拥有app: python-nginx-selector的pod 去访问指定ns下的pod 的指定端口
apiVersion: networking.k8s.io/v1 kind: NetworkPolicy metadata: name: egress-access-networkpolicy namespace: python spec: policyTypes: - Egress podSelector: #目标pod选择器 matchLabels: #基于label匹配目标pod app: python-nginx-selector #匹配python namespace中app的值为python-tomcat-app1-selector的pod,然后基于egress中的指定网络策略进行出口方向的网络限制 egress: - to: - namespaceSelector: matchLabels: nsname: python #指定允许访问的目的namespace - namespaceSelector: matchLabels: nsname: linux #指定允许访问的目的namespace ports: - protocol: TCP port: 8080 #允许80端口的访问 - protocol: TCP port: 53 #允许DNS的解析 - protocol: UDP port: 53