k8s集羣Canal的網絡控制

1 簡介

Canal 是calico和flannel的結合，我們因爲Calico不僅能夠提供很好的網絡性能，還能有網絡規則控制，但是我們很多時候使用的Flannel網絡，他的缺點就是沒有網絡控制只能提供網絡服務，所以Canal誕生了，爲Falanel提供了網絡控制。

網絡策略控制出站： egress 控制進站： ingress 可以以ip地址塊定義規則，也可以基於命名空間規則定義，還可以對pod定義規則

2 部署Canal

curl \
https://docs.projectcalico.org/v3.6/getting-started/kubernetes/installation/hosted/canal/canal.yaml \
-O
kubectl apply -f canal.yaml
1）canal默認控制的是10.244.0.0/16 ，如果你在安裝k8s集羣是設置的pod網絡不是此段網絡，則你需要修改配置文件
2）我們採用的和k8s集羣公用etcd的方式創建

查看創建的pod
# kubectl get pods -n kube-system 
NAME                                    READY   STATUS    RESTARTS   AGE
canal-9lmw8                             2/2     Running   0          4m50s
canal-ln6kg                             2/2     Running   0          4m50s
canal-mx6tn                             2/2     Running   0          4m50s
coredns-fb8b8dccf-lfbkh                 1/1     Running   2          10d
coredns-fb8b8dccf-t2kdz                 1/1     Running   2          10d
etcd-master-1                           1/1     Running   2          10d
kube-apiserver-master-1                 1/1     Running   2          10d
kube-controller-manager-master-1        1/1     Running   2          10d
kube-flannel-ds-amd64-df7gk             1/1     Running   3          9d
kube-flannel-ds-amd64-dzxfd             1/1     Running   3          9d
kube-flannel-ds-amd64-mgw2m             1/1     Running   2          9d
kube-proxy-47d6q                        1/1     Running   2          10d
kube-proxy-jztrs                        1/1     Running   3          10d
kube-proxy-rt4xx                        1/1     Running   3          10d
kube-scheduler-master-1                 1/1     Running   2          10d
kubernetes-dashboard-5f7b999d65-8h79h   1/1     Running   0          21h

3 創建兩個實驗的空間

kubectl create namespace shengchan
kubectl create namespace test

4 創建網絡規則測試

默認ingress 和 engress可以一起使用，當這兩規則同時存在那麼將使用policyTypes做具體的規則指向
例如：創建一個policy默認拒絕所有在空間shengchan的入站規則


# cat ingress-deny.yaml
apiVersion: networking.k8s.io/v1
kind: NetworkPolicy
metadata:
  name: deny-all-ingress
  namespace: shengchan
spec: 
  podSelector: {}
  policyTypes:
  - Ingress


kubectl apply ingress-deny.yaml

創建一組屬於shengchan空間的pod
# cat pod-networkpolicy1.yaml 
apiVersion: v1
kind: Pod
metadata:
   name: pod-test
spec:
 containers:
 - name: pod-test-policy
   image: nginx:1.10

kubectl apply -f pod-networkpolicy1.yaml  -n shengchan

創建一組屬於test空間的pod
kubectl apply -f pod-networkpolicy1.yaml  -n test

分別查看兩組pod的IP地址

# kubectl get pods -o wide  -n shengchan
NAME       READY   STATUS    RESTARTS   AGE   IP           NODE    NOMINATED NODE   READINESS GATES
pod-test   1/1     Running   0          16s   10.244.1.2   node2   <none>           <none>

[root@master-1 ~]# kubectl get pods -o wide  -n test
NAME       READY   STATUS    RESTARTS   AGE   IP           NODE    NOMINATED NODE   READINESS GATES
pod-test   1/1     Running   0          9s    10.244.2.2   node1   <none>           <none>


在宿主機上面分別ping這兩個ip
# ping 10.244.1.2 -c 1
PING 10.244.1.2 (10.244.1.2) 56(84) bytes of data.
^C
--- 10.244.1.2 ping statistics ---
1 packets transmitted, 0 received, 100% packet loss, time 0ms

# ping 10.244.2.2 -c 1
PING 10.244.2.2 (10.244.2.2) 56(84) bytes of data.
64 bytes from 10.244.2.2: icmp_seq=1 ttl=63 time=0.660 ms

--- 10.244.2.2 ping statistics ---
1 packets transmitted, 1 received, 0% packet loss, time 0ms
rtt min/avg/max/mdev = 0.660/0.660/0.660/0.000 ms

此時發現屬於shengchan空間的pod是被拒絕訪問的，屬於test空間的pod正常訪問
此時我們是基於名稱空間級別的入站網絡訪問策略

5 將上面的默認拒絕策略放行

修改配置文件
apiVersion: networking.k8s.io/v1
kind: NetworkPolicy
metadata:
  name: deny-all-ingress
  namespace: shengchan
spec: 
  podSelector: {}
  ingress:
  - {}
  policyTypes:
  - Ingress
kubectl apply -f ingress-deny.yaml

再次測試10.244.1.2

[root@master-1 ~]# curl -I 10.244.1.2
HTTP/1.1 200 OK
Server: nginx/1.10.3
Date: Sat, 27 Apr 2019 16:58:04 GMT
Content-Type: text/html
Content-Length: 612
Last-Modified: Tue, 31 Jan 2017 15:01:11 GMT
Connection: keep-alive
ETag: "5890a6b7-264"
Accept-Ranges: bytes

[root@master-1 ~]# curl -I 10.244.2.2
HTTP/1.1 200 OK
Server: nginx/1.10.3
Date: Sat, 27 Apr 2019 16:58:07 GMT
Content-Type: text/html
Content-Length: 612
Last-Modified: Tue, 31 Jan 2017 15:01:11 GMT
Connection: keep-alive
ETag: "5890a6b7-264"
Accept-Ranges: bytes

說明我們的網絡策略放行成功

6 放行特定的入站

以標籤的形式進行劃分pod入站規則是否被放行

1）回覆默認拒絕所有
[root@master-1 ~]# cat ingress-deny.yaml 
apiVersion: networking.k8s.io/v1
kind: NetworkPolicy
metadata:
  name: deny-all-ingress
  namespace: shengchan
spec: 
  podSelector: {}
  policyTypes:
  - Ingress

kubectl apply -f ingress-deny.yaml

2)給shengchan pod打一個標籤
kubectl label pods pod-test app=hello -n shengchan
3）創建一個新的規則
# vim ingress-allow.yml 

apiVersion: networking.k8s.io/v1
kind: NetworkPolicy
metadata:
  name: allow-hello-ingress
  namespace: shengchan
spec:
  podSelector:
    matchLabels:
      app: hello
  ingress:
  - from:
    - ipBlock:
        cidr: 10.244.0.0/16
        except:
        - 10.244.2.2/32
    ports:
    - protocol: TCP
      port: 80


解釋： 我們重新定義了一個policy，通過標籤匹配到我們已有的pod（我們一已有的pod是拒絕所有訪問的），定義ingres規則，如果不通過from方式定義默認是所有網段都可以訪問到擁有這個標籤的pod爲了測試我們使用from方式去拒絕10.244.0.0但是會放行10.244.2.2因爲我們有這個ip的pod創建萬郴更可以進行測試ports就是指定特定開放的端口
4）創建之前進行訪問測試
# kubectl get pods -n shengchan -o wide
NAME       READY   STATUS    RESTARTS   AGE   IP           NODE    NOMINATED NODE   READINESS GATES
pod-test   1/1     Running   0          62m   10.244.1.2   node2   <none>           <none>
# curl 10.244.1.2
說明無法訪問
5）創建之後測試
kubectl apply -f ingress-allow.yml 

[root@master-1 ~]# curl -I 10.244.1.2
HTTP/1.1 200 OK
Server: nginx/1.10.3
Date: Sat, 27 Apr 2019 17:32:15 GMT
Content-Type: text/html
Content-Length: 612
Last-Modified: Tue, 31 Jan 2017 15:01:11 GMT
Connection: keep-alive
ETag: "5890a6b7-264"
Accept-Ranges: bytes
說明已經放行
至於對於10.244.0.0網段的限制有興趣可以測試一下

6 egress出站規則

1）設置一個拒絕所有出站policy
cat egress.yml
apiVersion: networking.k8s.io/v1
kind: NetworkPolicy
metadata:
  name: deny-all-egress
  namespace: test
spec: 
  podSelector: {}
  policyTypes:
  - Egress

kubectl apply -f egress.yml

2）創建一個test命名空間下的pods
kubectl apply -f pod-networkpolicy1.yaml -n test
3)查看podIP並進去容器去測試ping一個正常的podip
[root@master-1 ~]# kubectl get pods -o wide
NAME                        READY   STATUS    RESTARTS   AGE    IP            NODE    NOMINATED NODE   READINESS GATES
net-test-5764c456cb-2c9df   1/1     Running   24         24h    10.244.2.22   node1   <none>           <none>
net-test-5764c456cb-ng6vh   1/1     Running   46         6d5h   10.244.1.21   node2   <none>           <none>

[root@master-1 ~]# kubectl get pods -o wide -n test
NAME       READY   STATUS    RESTARTS   AGE   IP           NODE    NOMINATED NODE   READINESS GATES
pod-test   1/1     Running   0          86m   10.244.2.2   node1   <none>           <none>

kubectl exec -it pod-test -n test -- /bin/bash

root@pod-test:/# ping 10.244.1.21
PING 10.244.1.21 (10.244.1.21): 56 data bytes
^C--- 10.244.1.21 ping statistics ---
3 packets transmitted, 0 packets received, 100% packet loss

說明此時已經應用策略
4）修改配置文件放行所有出站
# cat egress.yml 
apiVersion: networking.k8s.io/v1
kind: NetworkPolicy
metadata:
  name: deny-all-egress
  namespace: test
spec: 
  podSelector: {}
  egress:
  - {}
  policyTypes:
  - Egress

kubectl apply -f egress.yml 
測試：
root@pod-test:/# ping 10.244.1.21 -c 1
PING 10.244.1.21 (10.244.1.21): 56 data bytes
64 bytes from 10.244.1.21: icmp_seq=0 ttl=62 time=0.939 ms
--- 10.244.1.21 ping statistics ---
1 packets transmitted, 1 packets received, 0% packet loss
round-trip min/avg/max/stddev = 0.939/0.939/0.939/0.000 ms
此時說明已經放行成功

7 更加複雜的policy規則需要根據不同的環境規劃進行深入探究，其所有都規則都體現在配置文件中

k8s集羣Canal的網絡控制

DAPPER 事務 TRANSACTION

深圳IO-被動紅外感應設計

深圳IO-發光標誌

深圳IO-飲酒遊戲計分器

深圳IO-脈衝發生器

Java中線程的創建方式

https://yachay.unat.edu.pe/blog/index.php?comment_area=format_blog&comment_component=blog&comment_co

linux以太網驅動總結