概念

Service可以看作是一組提供相同服務的Pod對外的訪問接口。藉助Service,應
用可以方便地實現服務發現和負載均衡。
service默認只支持4層負載均衡能力,沒有7層功能。(可以通過Ingress實現)
service的類型:
- ClusterIP:默認值,k8s系統給service自動分配的虛擬IP,只能在集羣內部訪問。
- NodePort:將Service通過指定的Node上的端口暴露給外部,訪問任意一個
  NodeIP:nodePort都將路由到ClusterIP。
- LoadBalancer:在 NodePort 的基礎上,藉助 cloud provider 創建一個外部的負
  載均衡器,並將請求轉發到 :NodePort,此模式只能在雲服務器上使用。
- ExternalName:將服務通過 DNS CNAME 記錄方式轉發到指定的域名(通過
  spec.externlName 設定)。

我們不設置service的類型時，默認爲clusterIP類型;

service

ClusterIP

[root@server2 manifest]# cat service.yml 
kind: Service
apiVersion: v1
metadata:
  name: myservice
spec:
  ports:		//這裏並沒有指定service類型
    - protocol: TCP
      port: 80
      targetPort: 80
  selector:
    app: myapp
[root@server2 manifest]# kubectl apply -f service.yml 
kuservice/myservice created
[root@server2 manifest]# kubectl describe svc myservice 
Name:              myservice
Namespace:         default
Labels:            <none>
Annotations:       Selector:  app=myapp
Type:              ClusterIP		//默認clusterIP類型

NodePort

[root@server2 manifest]# kubectl apply -f pod2.yml 
deployment.apps/deployment-example created
[root@server2 manifest]# kubectl describe svc myservice 
Name:              myservice
Namespace:         default
Labels:            <none>
Annotations:       Selector:  app=myapp
Type:              ClusterIP
IP:                10.108.247.199
Port:              <unset>  80/TCP
TargetPort:        80/TCP
Endpoints:         10.244.1.83:80,10.244.2.58:80		//加進去了
Session Affinity:  None
Events:            <none>  
[root@server2 manifest]# kubectl edit svc myservice 	//編輯svc
    app: myapp
  sessionAffinity: None
  type: NodePOrt	//這裏原本是ClusterIP

[root@server2 manifest]# kubectl get svc
NAME         TYPE        CLUSTER-IP       EXTERNAL-IP   PORT(S)        AGE
kubernetes   ClusterIP   10.96.0.1        <none>        443/TCP        8d
myservice    NodePort    10.108.247.199   <none>        80:31972/TCP   5m44s	//31972端口打開
[root@server2 manifest]# kubectl get pod -o wide
NAME                                  READY   STATUS    RESTARTS   AGE     IP            NODE      NOMINATED NODE   READINESS GATES
deployment-example-5c9fb4c54c-jrb88   1/1     Running   0          2m24s   10.244.1.83   server3   <none>           <none>
deployment-example-5c9fb4c54c-q95v6   1/1     Running   0          2m24s   10.244.2.58   server4   <none>           <none>

//在server3和4上查看端口
[root@server3 ~]# netstat -tnlp|grep 31972
tcp        0      0 0.0.0.0:31972           0.0.0.0:*               LISTEN      4020/kube-proxy
[root@server4 ~]# netstat -tnlp|grep 31972
tcp        0      0 0.0.0.0:31972           0.0.0.0:*               LISTEN      4020/kube-proxy

都打開了我們就可以去集羣外部訪問了。

ipvs模式的service

Service 是由 kube-proxy 組件,加上 iptables 來共同實現的.
kube-proxy 通過 iptables 處理 Service 的過程,需要在宿主機上設置相當多的
iptables 規則,如果宿主機有大量的Pod,不斷刷新iptables規則,會消耗大量的
CPU資源。

[root@server3 ~]# iptables -t nat -nL

service在各個結點上運用了大量的nat規則。ipvs模式的service可以讓k8s支持更大量級的pod。
開啓kube-proxy的ipvs模式:

[root@server2 manifest]# yum install -y ipvsadm
[root@server3 ~]# yum install -y ipvsadm
[root@server4 ~]# yum install -y ipvsadm
//所有結點安裝ipvsadm
[root@server2 manifest]# lsmod |grep ip_vs
ip_vs_sh               12688  0 
ip_vs_wrr              12697  0 
ip_vs_rr               12600  0 
ip_vs                 145497  6 ip_vs_rr,ip_vs_sh,ip_vs_wrr
nf_conntrack          133095  9 ip_vs,nf_nat,nf_nat_ipv4,nf_nat_ipv6,xt_conntrack,nf_nat_masquerade_ipv4,nf_conntrack_netlink,nf_conntrack_ipv4,nf_conntrack_ipv6
libcrc32c              12644  4 xfs,ip_vs,nf_nat,nf_conntrack

修改ipvs的模式

[root@server2 manifest]# kubectl edit cm kube-proxy -n kube-system
    mode: "ipvs"		//指定使用ipvs模式

[root@server2 manifest]# kubectl get pod -n kube-system |grep kube-proxy | awk '{system("kubectl delete pod "$1" -n kube-system")}'
pod "kube-proxy-7kn82" deleted
pod "kube-proxy-hww5t" deleted
pod "kube-proxy-wn4h8" deleted
//更新kube-proxy pod，刪除後回進行重啓策略。加載更改內容。

[root@server2 manifest]# ipvsadm -ln
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
TCP  10.96.0.1:443 rr
  -> 172.25.254.2:6443            Masq    1      0          0         
TCP  10.96.0.10:53 rr
  -> 10.244.0.11:53               Masq    1      0          0         
  -> 10.244.0.12:53               Masq    1      0          0         
TCP  10.96.0.10:9153 rr
  -> 10.244.0.11:9153             Masq    1      0          0         
  -> 10.244.0.12:9153             Masq    1      0          0         
TCP  10.103.6.176:80 rr
  -> 10.244.1.83:80               Masq    1      0          0         
  -> 10.244.2.58:80               Masq    1      0          0         
UDP  10.96.0.10:53 rr
  -> 10.244.0.11:53               Masq    1      0          0         
  -> 10.244.0.12:53               Masq    1      0          0
//kube-proxy通過linux的IPVS模塊,以rr輪詢方式調度service中的Pod。

IPVS模式下,kube-proxy會在service創建後,在宿主機上添加一個虛擬網卡:
kube-ipvs0,並分配service IP:

[root@server2 manifest]# kubectl apply -f service.yml 
service/myservice created
[root@server2 manifest]# ip a

9: kube-ipvs0: <BROADCAST,NOARP> mtu 1500 qdisc noop state DOWN group default 
    link/ether b2:df:38:0a:cb:02 brd ff:ff:ff:ff:ff:ff
    inet 10.96.0.10/32 brd 10.96.0.10 scope global kube-ipvs0
       valid_lft forever preferred_lft forever
    inet 10.96.0.1/32 brd 10.96.0.1 scope global kube-ipvs0
       valid_lft forever preferred_lft forever
    inet `10.103.6.176/32` brd 10.103.6.176 scope global kube-ipvs0
       valid_lft forever preferred_lft forever
[root@server2 manifest]# kubectl describe svc myservice 
Name:              myservice
Namespace:         default
Labels:            <none>
Annotations:       Selector:  app=myapp
Type:              ClusterIP
IP:                `10.103.6.176`	
Port:              <unset>  80/TCP
TargetPort:        80/TCP
Endpoints:         10.244.1.83:80,10.244.2.58:80
Session Affinity:  None
Events:            <none>

並且每個結點上都有。我們就可以實現負載均衡了。

[root@server2 manifest]# kubectl run demo  --image=busyboxplus -it --restart=Never
If you don't see a command prompt, try pressing enter.
/ # curl 10.103.6.176:80
Hello MyApp | Version: v2 | <a href="hostname.html">Pod Name</a>
/ # curl 10.103.6.176:80/hostname.html
deployment-example-5c9fb4c54c-q95v6
/ # curl 10.103.6.176:80/hostname.html
deployment-example-5c9fb4c54c-jrb88
/ # curl 10.103.6.176:80/hostname.html
deployment-example-5c9fb4c54c-q95v6
/ # curl 10.103.6.176:80/hostname.html
deployment-example-5c9fb4c54c-jrb88

[root@server4 ~]# ipvsadm -ln
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
TCP  10.96.0.1:443 rr
  -> 172.25.254.2:6443            Masq    1      0          0         
TCP  10.96.0.10:53 rr
  -> 10.244.0.11:53               Masq    1      0          0         
  -> 10.244.0.12:53               Masq    1      0          0         
TCP  10.96.0.10:9153 rr
  -> 10.244.0.11:9153             Masq    1      0          0         
  -> 10.244.0.12:9153             Masq    1      0          0         
TCP  10.103.6.176:80 rr
  -> 10.244.1.83:80               Masq    1      0          3         
  -> 10.244.2.58:80               Masq    1      0          3        //這裏
UDP  10.96.0.10:53 rr
  -> 10.244.0.11:53               Masq    1      0          0         
  -> 10.244.0.12:53               Masq    1      0          0

擴容：

[root@server2 manifest]# vim pod2.yml
apiVersion: apps/v1
kind: Deployment
metadata:
  name: deployment-example
spec:
  replicas: 4	//改成4個

[root@server2 manifest]# kubectl apply -f pod2.yml 
deployment.apps/deployment-example configured
[root@server2 manifest]# kubectl get pod -o wide
NAME                                  READY   STATUS    RESTARTS   AGE    IP            NODE      NOMINATED NODE   READINESS GATES
deployment-example-5c9fb4c54c-jrb88   1/1     Running   0          125m   10.244.1.83   server3   <none>           <none>
deployment-example-5c9fb4c54c-pwjv5   1/1     Running   0          9s     10.244.2.63   server4   <none>           <none>
deployment-example-5c9fb4c54c-q95v6   1/1     Running   0          125m   10.244.2.58   server4   <none>           <none>
deployment-example-5c9fb4c54c-zkj2h   1/1     Running   0          8s     10.244.1.87   server3   <none>           <none>

[root@server4 ~]# ipvsadm -ln
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
TCP  10.96.0.1:443 rr
  -> 172.25.254.2:6443            Masq    1      0          0         
TCP  10.96.0.10:53 rr
  -> 10.244.0.11:53               Masq    1      0          0         
  -> 10.244.0.12:53               Masq    1      0          0         
TCP  10.96.0.10:9153 rr
  -> 10.244.0.11:9153             Masq    1      0          0         
  -> 10.244.0.12:9153             Masq    1      0          0         
TCP  10.103.6.176:80 rr
  -> 10.244.1.83:80               Masq    1      0          0         
  -> 10.244.1.87:80               Masq    1      0          0         
  -> 10.244.2.58:80               Masq    1      0          0         
  -> 10.244.2.63:80               Masq    1      0          0    //生效了     
UDP  10.96.0.10:53 rr
  -> 10.244.0.11:53               Masq    1      0          0         
  -> 10.244.0.12:53               Masq    1      0          0

這樣來看比iptables確實簡潔清爽的多。

Kubernetes 提供了一個 DNS 插件 Service：

[root@server2 manifest]# kubectl describe svc myservice 
Name:              myservice
Namespace:         default
Labels:            <none>
Annotations:       Selector:  app=myapp
Type:              ClusterIP
IP:                10.103.6.176

[root@server2 manifest]# kubectl run demo  --image=busyboxplus -it --restart=Never

/ # nslookup myservice.default.svc.cluster.local
Server:    10.96.0.10
Address 1: 10.96.0.10 kube-dns.kube-system.svc.cluster.local

Name:      myservice.default.svc.cluster.local
Address 1: 10.103.6.176 myservice.default.svc.cluster.local
// 會解析到service的ip
[root@server4 ~]# ipvsadm -ln
...  
TCP  10.103.6.176:80 rr
  -> 10.244.1.83:80               Masq    1      0          0         
  -> 10.244.1.87:80               Masq    1      0          0         
  -> 10.244.2.58:80               Masq    1      0          0         
  -> 10.244.2.63:80               Masq    1      0          0 


回收重啓：
[root@server2 manifest]# kubectl delete -f service.yml 
service "myservice" deleted
[root@server2 manifest]# kubectl apply -f service.yml 
service/myservice created
[root@server2 manifest]# kubectl describe svc myservice 
Name:              myservice
Namespace:         default
Labels:            <none>
Annotations:       Selector:  app=myapp
Type:              ClusterIP
IP:                10.109.129.13	//地址改變

[root@server4 ~]# ipvsadm -ln      
TCP  10.109.129.13:80 rr
  -> 10.244.1.83:80               Masq    1      0          0         
  -> 10.244.1.87:80               Masq    1      0          0         
  -> 10.244.2.58:80               Masq    1      0          0         
  -> 10.244.2.63:80               Masq    1      0          0 
結點上也會進行更新，整個集羣都會變化。

[root@server2 manifest]# kubectl run demo  --image=busyboxplus -it --restart=Never
If you don't see a command prompt, try pressing enter.
/ # nslookup myservice.default.svc.cluster.local
Server:    10.96.0.10
Address 1: 10.96.0.10 kube-dns.kube-system.svc.cluster.local

Name:      myservice.default.svc.cluster.local
Address 1: 10.109.129.13 myservice.default.svc.cluster.local		//解析也變了
/ # curl myservice.default.svc.cluster.local		//我們也可以直接使用域名進行訪問
Hello MyApp | Version: v2 | <a href="hostname.html">Pod Name</a>
/ # curl myservice.default.svc.cluster.local/hostname.html
deployment-example-5c9fb4c54c-zkj2h
/ # curl myservice.default.svc.cluster.local/hostname.html
deployment-example-5c9fb4c54c-jrb88
/ # curl myservice.default.svc.cluster.local/hostname.html
deployment-example-5c9fb4c54c-pwjv5
/ # curl myservice.default.svc.cluster.local/hostname.html
deployment-example-5c9fb4c54c-q95v6			//負載均衡

這樣的話vip改變我們也不用去改變了，直接訪問域名就可以自動解析，獲取服務了。
簡化：

/ # curl myservice
Hello MyApp | Version: v2 | <a href="hostname.html">Pod Name</a>
/ # curl myservice
Hello MyApp | Version: v2 | <a href="hostname.html">Pod Name</a>
/ # nslookup myservice
Server:    10.96.0.10
Address 1: 10.96.0.10 kube-dns.kube-system.svc.cluster.local

Name:      myservice
Address 1: 10.109.129.13 myservice.default.svc.cluster.local	//默認會解析到後面
/ # curl myservice/hostname.html
deployment-example-5c9fb4c54c-pwjv5
/ # curl myservice/hostname.html
deployment-example-5c9fb4c54c-q95v6
/ # curl myservice/hostname.html
deployment-example-5c9fb4c54c-zkj2h
/ # curl myservice/hostname.html
deployment-example-5c9fb4c54c-jrb88

直接這樣訪問就行了。

我們還可以使用nodeport的模式：

[root@server2 manifest]# kubectl edit svc myservice
  type: NodePort
[root@server2 manifest]# kubectl get svc
NAME         TYPE        CLUSTER-IP      EXTERNAL-IP   PORT(S)        AGE
kubernetes   ClusterIP   10.96.0.1       <none>        443/TCP        9d
myservice    NodePort    10.109.129.13   <none>        80:31635/TCP   17h  /暴露端口

[root@server3 ~]# ipvsadm -ln |grep 31635
TCP  10.244.1.0:31635 rr
TCP  127.0.0.1:31635 rr
TCP  172.17.0.1:31635 rr
TCP  172.25.254.3:31635 rr
TCP  10.244.1.1:31635 rr
/ 其它的結點上也暴露了端口，我們可以在集羣外部進行訪問了。

root@rhel7host ~]# curl 172.25.254.3:31635/hostname.html
deployment-example-5c9fb4c54c-srkz9
[root@rhel7host ~]# curl 172.25.254.3:31635/hostname.html
deployment-example-5c9fb4c54c-kpwwm
[root@rhel7host ~]# curl 172.25.254.3:31635/hostname.html
deployment-example-5c9fb4c54c-ljsjf

service的"無頭服務"

Headless Service “無頭服務”
Headless Service不需要分配一個VIP,而是直接以DNS記錄的方式解析出被代理
Pod的IP地址。
域名格式: $(servicename).$ (namespace).svc.cluster.local

我們最好安裝bing-utils工具：
bind是linux系統下的一個DNS服務程序.bind-utils是bind軟件提供的一組DNS工具包,
裏面有一些DNS相關的工具.主要有:dig,host,nslookup,nsupdate.使用這些工具可以
進行域名解析和DNS調試工作.

[root@server2 manifest]# yum install bind-utils -y

然後我們就可以使用dig命令：

[root@server2 manifest]# kubectl get svc
NAME         TYPE        CLUSTER-IP      EXTERNAL-IP   PORT(S)        AGE
kubernetes   ClusterIP   10.96.0.1       <none>        443/TCP        9d
myservice    NodePort    10.109.129.13   <none>        80:31635/TCP   17h	/myservice地址

[root@server2 manifest]# kubectl get svc -n kube-system
NAME       TYPE        CLUSTER-IP   EXTERNAL-IP   PORT(S)                  AGE
kube-dns   ClusterIP   10.96.0.10   <none>        53/UDP,53/TCP,9153/TCP   9d	/dns地址

[root@server2 manifest]# dig myservice.default.svc.cluster.local @10.96.0.1		/@後接dns地址
;; ANSWER SECTION:
myservice.default.svc.cluster.local. 30	IN A	10.109.129.13			/解析到了

我們當前有兩個dns服務：

[root@server2 manifest]# kubectl get pod -n kube-system -o wide
NAME                              READY   STATUS    RESTARTS   AGE   IP             NODE      NOMINATED NODE   READINESS GATES
coredns-5fd54d7f56-k5jb9          1/1     Running   5          9d    10.244.0.13    server2   <none>           <none>
coredns-5fd54d7f56-xb5m4          1/1     Running   6          9d    10.244.0.14    server2   <none>           <none>

當我們的訪問量過大時，我們還可以對dns服務pod進行拉伸，因爲它是以deployment控制器控制的：

[root@server2 manifest]# kubectl -n kube-system get all
NAME                                  READY   STATUS    RESTARTS   AGE
...
NAME                      READY   UP-TO-DATE   AVAILABLE   AGE
deployment.apps/coredns   2/2     2            2           9d

NAME                                 DESIRED   CURRENT   READY   AGE
replicaset.apps/coredns-5fd54d7f56   2         2         2       9d

那我們的無頭服務怎麼設置哪？

[root@server2 manifest]# vim service.yml 
[root@server2 manifest]# cat service.yml 
kind: Service
apiVersion: v1
metadata:
  name: myservice
spec:
  ports:
    - protocol: TCP
      port: 80
      targetPort: 80
  selector:
    app: myapp
  clusterIP: None		//不設置clusterip

[root@server2 manifest]# kubectl describe svc myservice 
Name:              myservice
Namespace:         default
Labels:            <none>
Annotations:       Selector:  app=myapp
Type:              ClusterIP
IP:                None
Port:              <unset>  80/TCP
TargetPort:        80/TCP
Endpoints:         10.244.1.91:80,10.244.1.92:80,10.244.2.67:80 + 1 more...
Session Affinity:  None
Events:            <none>

		這個服務就沒有分配到IP了,但是後端仍然存在

那我們怎樣去訪問哪？通過dns解析訪問域名就行了。

[root@server2 manifest]# dig myservice.default.svc.cluster.local. @10.96.0.10
;; ANSWER SECTION:
myservice.default.svc.cluster.local. 30	IN A	10.244.1.91
myservice.default.svc.cluster.local. 30	IN A	10.244.1.92
myservice.default.svc.cluster.local. 30	IN A	10.244.2.67
myservice.default.svc.cluster.local. 30	IN A	10.244.2.68
//正好對應了四個後端的IP

滾動更新後解析到的地址也會改變：

loadbalancer類型的svc

從外部訪問 Service 的第二種方式,適用於公有云上的 Kubernetes 服務。這
時候,你可以指定一個 LoadBalancer 類型的 Service。

[root@server2 manifest]# vim service.yml 
kind: Service
apiVersion: v1
metadata:
  name: myservice
spec:
  ports:
    - protocol: TCP
      port: 80
      targetPort: 80
  selector:
    app: myapp
  type: LoadBalancer		/設置爲lodabalancer類型
[root@server2 manifest]# kubectl apply -f service.yml 
service/myservice created
[root@server2 manifest]# kubectl get svc
NAME         TYPE           CLUSTER-IP       EXTERNAL-IP   PORT(S)        AGE
kubernetes   ClusterIP      10.96.0.1        <none>        443/TCP        9d
myservice    LoadBalancer   10.103.172.225   <pending>     80:32752/TCP   5s
										/pending是正在調用外部共有云

在service提交後,Kubernetes就會調用 CloudProvider 在公有云上爲你創建一個負載均衡服務,並且把被代理的 Pod 的 IP地址配置給負載均衡服務做後端。

ExternalName

外部訪問的第三種方式叫做ExternalName。在及集羣中也可能需要去訪問外部的網站，但是外部的資源是不可控的，當外部域名改變時，我們想要不影響我們的集羣內部的訪問，我們就需要指定一個ExternalName類型的service，當域名改變時，我們只在service中更改就行了，後端的訪問效果不變。

[root@server2 manifest]# vim service.yml
kind: Service
apiVersion: v1
metadata:
  name: myservice
spec:
  ports:
    - protocol: TCP
      port: 80
      targetPort: 80
  selector:
    app: myapp
  type: ExternalName
  externalName: www.baidu.com		/指定一個域名

[root@server2 manifest]# kubectl apply -f  service.yml 
service/myservice created
[root@server2 manifest]# kubectl get svc myservice 
NAME        TYPE           CLUSTER-IP   EXTERNAL-IP     PORT(S)   AGE
myservice   ExternalName   <none>       www.baidu.com   80/TCP    13s
[root@server2 manifest]# dig myservice.default.svc.cluster.local. @10.244.0.13

;; ANSWER SECTION:
myservice.default.svc.cluster.local. 30	IN CNAME www.baidu.com.
www.baidu.com.		30	IN	CNAME	www.a.shifen.com.
www.a.shifen.com.	30	IN	A	36.152.44.95
www.a.shifen.com.	30	IN	A	36.152.44.96		//解析到了

我們變更一下域名。

[root@server2 manifest]# vim service.yml 
  externalName: www.baidu.com		//改變
[root@server2 manifest]# kubectl apply -f service.yml 
service/myservice configured

[root@server2 manifest]# kubectl get svc myservice 
NAME        TYPE           CLUSTER-IP   EXTERNAL-IP   PORT(S)   AGE
myservice   ExternalName   <none>       www.qq.com    80/TCP    110s
[root@server2 manifest]# dig myservice.default.svc.cluster.local. @10.244.0.13


;; ANSWER SECTION:
myservice.default.svc.cluster.local. 30	IN CNAME www.qq.com.
www.qq.com.		30	IN	CNAME	public-v6.sparta.mig.tencent-cloud.net.
public-v6.sparta.mig.tencent-cloud.net.	30 IN A	117.184.242.202
public-v6.sparta.mig.tencent-cloud.net.	30 IN A	183.192.170.139
public-v6.sparta.mig.tencent-cloud.net.	30 IN A	183.192.170.170

同樣也解析到了，我們訪問的地址並沒有改變，一直都是myservice.default.svc.cluster.local.，但是我們能解析到不同的域名。

爲service分配一個公有IP

[root@server2 manifest]# vim service.yml 
kind: Service
apiVersion: v1
metadata:
  name: myservice
spec:
  ports:
    - protocol: TCP
      port: 80
      targetPort: 80
  selector:
    app: myapp
  externalIPs:
    - 172.25.254.100		//指定一個外部的合法IP

[root@server2 manifest]# kubectl apply -f service.yml 
service/myservice created
[root@server2 manifest]# kubectl get svc
NAME         TYPE        CLUSTER-IP       EXTERNAL-IP    PORT(S)   AGE
kubernetes   ClusterIP   10.96.0.1        <none>         443/TCP   9d
myservice    ClusterIP   10.110.224.207   172.25.0.100   80/TCP    11s
// 它會把這個ip直接和service的ip綁定

[root@server2 manifest]# ipvsadm -ln
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
TCP  172.25.254.100:80 rr		//也綁定到了ipvs上
  -> 10.244.1.93:80               Masq    1      0          0         
  -> 10.244.1.94:80               Masq    1      0          0         
  -> 10.244.1.95:80               Masq    1      0          0

[root@server1 harbor]# curl 172.25.254.100		/可以直接訪問了
Hello MyApp | Version: v1 | <a href="hostname.html">Pod Name</a>

[root@server3 ~]# ipvsadm -ln 
IP Virtual Server version 1.2.1 (size=4096)
Prot LocalAddress:Port Scheduler Flags
  -> RemoteAddress:Port           Forward Weight ActiveConn InActConn
TCP  172.25.254.100:80 rr
  -> 10.244.1.93:80               Masq    1      0          4         
  -> 10.244.1.94:80               Masq    1      0          4         
  -> 10.244.1.95:80               Masq    1      0          4         
  -> 10.244.2.69:80               Masq    1      0          5

Kubernetes進階 -- service不同模式理解與實踐

概念

service

ClusterIP

NodePort

ipvs模式的service

service的"無頭服務"

loadbalancer類型的svc

ExternalName

爲service分配一個公有IP

認知提升的方法

螞蟻面試：Springcloud核心組件的底層原理，你知道多少？

C#開源的兩款功能強大的錄屏神器

kubernetes -- 結點調度控制的幾種方式

kubernetes存儲 -- Volumes管理（三）StatefulSet控制器、StatefulSet部署mysql主從集羣

kubernetes存儲 -- Volumes管理（一）emptyDir卷、hostpath卷、NFS卷

kubernetes存儲 -- Volumes管理（二）pv持久卷 +NFS、存儲類。

kubernetes存儲 -- Secret配置管理

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