@Pod资源共享机制
Kubernetes直接管理Pod，而不是容器
Pod不是进程，而是容器运行的环境
Pod是Kubernetes创建和管理的最小单元，不同Pod可以在不同Node上
Pod中可以有一个或多个容器[1]，且始终部署在一个Node上，这些容器共享存储、网络
说明[1]:
Pod中运行单个容器 Pod可以看作是单个容器的抽象封装
Pod中运行多个容器 需要共享资源或紧密耦合的应用程序[2],Pod 将这些容器和存储资源打包为一个可管理的实体。
说明[2]:
应用程序之间发生文件、数据交互
应用程序之间需要通过本地通信(127.0.0.1、socket)
应用程序之间频繁调用
官方文档传送门
Pods -> https://kubernetes.io/zh/docs/concepts/workloads/pods/
Init容器 -> https://kubernetes.io/zh/docs/concepts/workloads/pods/init-containers
在每个Node上执行docker ps 会发现有不少pause
pause容器/infra (infrastucture container)容器和业务容器一一对应，业务容器共享pause容器的网络栈和Volume挂载卷。
[root@k8s-master ~]# docker ps | grep dash
64cea4b64eb1 86262685d9ab "/metrics-sidecar" 3 hours ago Up 3 hours k8s_dashboard-metrics-scraper_dashboard-metrics-scraper-7b59f7d4df-xqb96_kubernetes-dashboard_7435c1e5-359b-4350-8fa7-44facdc33e63_8
fc7a1d65ad43 registry.aliyuncs.com/google_containers/pause:3.2 "/pause" 3 hours ago Up 3 hours k8s_POD_dashboard-metrics-scraper-7b59f7d4df-xqb96_kubernetes-dashboard_7435c1e5-359b-4350-8fa7-44facdc33e63_8
[root@k8s-master ~]#
[root@k8s-master ~]# docker ps | grep etcd
4f0eadebcb08 0369cf4303ff "etcd --advertise-cl…" 3 hours ago Up 3 hours k8s_etcd_etcd-k8s-master_kube-system_729d533a85352265aebac662f2565332_12
e5d19e95ec85 registry.aliyuncs.com/google_containers/pause:3.2 "/pause" 3 hours ago Up 3 hours k8s_POD_etcd-k8s-master_kube-system_729d533a85352265aebac662f2565332_12
[root@k8s-master ~]#
[root@k8s-node1 ~]# docker ps | grep web
a049aebee4bb nginx "/docker-entrypoint.…" 25 minutes ago Up 25 minutes k8s_nginx_web-6f7dfb48c4-gmr8d_default_8ed106ee-b02e-4f26-a7c4-c4f014e4c192_0
0330feccf8b4 registry.aliyuncs.com/google_containers/pause:3.2 "/pause" 26 minutes ago Up 26 minutes k8s_POD_web-6f7dfb48c4-gmr8d_default_8ed106ee-b02e-4f26-a7c4-c4f014e4c192_0
[root@k8s-node1 ~]#@Pod资源共享举例(网络)
创建一个包含两个容器的Pod
注意：单独创建Pod可以使用命令例如 kubectl run nginx --image=nginx 或 yaml文件，这里举例使用yaml文件。一般不单独创建Pod，而是通过控制器创建。
#清理deployment
[root@k8s-master ~]# kubectl delete deploy --all
deployment.apps "web" deleted
[root@k8s-master ~]#
#编辑Pod的yaml文件
[root@k8s-master ~]# cat test-netshare.yaml
apiVersion: v1
kind: Pod
metadata:
labels:
app: test
name: test-netshare
namespace: default
spec:
containers:
- image: busybox
name: hello
command: ["/bin/sh","-c",'echo "Hello, Kubernetes!" && sleep 3600']
- image: nginx
name: web
[root@k8s-master ~]#
#创建Pod
[root@k8s-master ~]# kubectl apply -f test-netshare.yaml
pod/test-netshare created
[root@k8s-master ~]#kubectl get pod -o wide可以看到Pod的IP
kubectl describe pod <pod name>，可以看到test-netshare这个Pod里有两个容器，hello和web
[root@k8s-master ~]# kubectl get pod -o wide
NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES
test-netshare 2/2 Running 0 2m49s 10.244.36.107 k8s-node1 <none> <none>
[root@k8s-master ~]#
[root@k8s-master ~]# kubectl describe pod test-netshare
Name: test-netshare
……
IPs:
IP: 10.244.36.107
Containers:
hello:
Container ID: docker://73b6704f226233c1776b1fc4ab66a1e2b6326b108f77e5dc1d57b338b68193d8
Image: busybox
……
web:
Container ID: docker://459f0e2bf7bc33f66dfc523ec3210229783c86ddb27fde309224b481a3219502
Image: nginx
……首先进入容器hello看看，注意ipaddr输出的结果，和上述分配给Pod的IP一致
[root@k8s-master ~]# kubectl exec -it test-netshare -c hello -- sh
/ # ip addr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
2: tunl0@NONE: <NOARP> mtu 1480 qdisc noop qlen 1000
link/ipip 0.0.0.0 brd 0.0.0.0
4: eth0@if19: <BROADCAST,MULTICAST,UP,LOWER_UP,M-DOWN> mtu 1440 qdisc noqueue
link/ether 96:02:7f:2d:9c:f0 brd ff:ff:ff:ff:ff:ff
inet 10.244.36.107/32 scope global eth0
valid_lft forever preferred_lft forever
/ #由于另一个容器web用的nginx镜像，这里wget一把，能够下载到nginx的主页
[root@k8s-master ~]# kubectl exec -it test-netshare -c hello -- sh
/ # curl 127.0.0.1:80
sh: curl: not found
/ # wget 127.0.0.1:80
Connecting to 127.0.0.1:80 (127.0.0.1:80)
saving to 'index.html'
index.html 100% |*********************************************************************************************************| 612 0:00:00 ETA
'index.html' saved
/ # cat index.html
<!DOCTYPE html>
<html>
<head>
<title>Welcome to nginx!</title>
……
</html>
/ #
说明：web容器不支持查看网络的命令
[root@k8s-master ~]# kubectl exec -it test-netshare -c web -- sh
#
# ip addr
sh: 2: ip: not found
# ifconfig
sh: 3: ifconfig: not found
# cd /etc/sysconfig
sh: 4: cd: can't cd to /etc/sysconfig@Pod资源共享举例(存储)
创建一个包含两个容器的Pod
[root@k8s-master ~]# cat test-volshare.yaml
apiVersion: v1
kind: Pod
metadata:
labels:
app: test
name: test-volshare
namespace: default
spec:
containers:
- image: busybox
name: hello
command: ["/bin/sh","-c",'echo "Hello, Kubernetes!" && sleep 3600']
volumeMounts:
- name: log
mountPath: /var/data
- image: nginx
name: web
volumeMounts:
- name: log
mountPath: /usr/share/data
volumes:
- name: log
emptyDir: {}
[root@k8s-master ~]#
[root@k8s-master ~]# kubectl apply -f test-volshare.yaml
pod/test-volshare created
[root@k8s-master ~]#先进入容器hello，在其数据卷挂载的路径下创建一个文件
[root@k8s-master ~]# kubectl get pod
NAME READY STATUS RESTARTS AGE
test-volshare 2/2 Running 0 67s
[root@k8s-master ~]#
[root@k8s-master ~]# kubectl exec -it test-volshare -c hello -- sh
/ # cd /var/data
/var/data # ls
/var/data # touch createbyhello
/var/data # echo 'hello' >> createbyhello
/var/data # exit
[root@k8s-master ~]#然后进入容器web，在其数据卷挂载的路径下可以看到刚才在hello容器中创建的文件
[root@k8s-master ~]# kubectl exec -it test-volshare -c web -- bash
root@test-volshare:/# cd /usr/share/data/
root@test-volshare:/usr/share/data# ls
createbyhello
root@test-volshare:/usr/share/data# cat createbyhello
hello
root@test-volshare:/usr/share/data#