基于CentOS7.2安装Kubernetes-v1.2

摘要

使用swarm构建docker集群之后我们发现面临很多问题 swarm虽好但是还处于发展阶段功能上有所不足 我们使用kubernetes来解决这个问题

kubernetes 与swarm 比较

优点

• 复制集与健康维护

• 服务自发现与负载均衡

• 灰度升级

• 垃圾回收 自动回收失效镜像与容器

• 与容器引擎解耦 不仅仅支持docker容器

• 用户认证与资源隔离

缺点

大而全意味着 复杂度较高 从部署到使用都比swarm 复杂的多 相对而已swarm比较轻量级 而且跟docker引擎配合的更好 从精神上我是更支持swarm 奈何现在功能太欠缺 几天前发布了一个 SwarmKit的管理功能 功能多了不少 待其完善以后可以重回swarm的怀抱

K8s 核心概念简介

pod

k8s 中创建的最小部署单元就是pod 而容器是运行在pod里面的 pod可以运行多个容器 pod内的容器可以共享网络和存储相互访问

Replication controller

复制及控制器：对多个pod创建相同的副本运行在不同节点 一般不会创建单独的pod而是与rc配合创建 rc控制并管理pod的生命周期维护pod的健康

service

每个容器重新运行后的ip地址都不是固定的 所以要有一个服务方向和负载均衡来处理 service就可以实现这个需求 service创建后可以暴露一个固定的端口 与相应的pod 进行绑定

K8s 核心组件简介

apiserver

@H_618_134@提供对外的REST API服务 运行在 master节点 对指令进行验证后 修改etcd的存储数据

@H_618_134@shcheduler

@H_618_134@调度器运行在master节点，通过apiserver定时监控数据变化 运行pod时通过自身调度算法选出可运行的节点

@H_618_134@controller-manager

@H_618_134@控制管理器运行在master节点 分别几大管理器定时运行 分别为

@H_618_134@1）Replication controller 管理器 管理并保存所有的rc的的状态

@H_618_134@2 ) service Endpoint 管理器 对service 绑定的pod 进行实时更新操作 对状态失败的pod进行解绑

@H_618_134@3）Node controller 管理器 定时对集群的节点健康检查与监控

@H_618_134@4）资源配额管理器 追踪集群资源使用情况

@H_618_134@kuctrl （子节点）

@H_618_134@管理维护当前子节点的所有容器 如同步创建新容器 回收镜像垃圾等

@H_618_134@kube-proxy （子节点）

对客户端请求进行负载均衡并分配到service后端的pod 是service的具体实现保证了ip的动态变化 proxy 通过修改iptable 实现路由转发

工作流程

k8s 安装过程：

一、主机规划表:

IP地址	@H_867_197@角色	安装软件包	@H_867_197@启动服务及顺序
192.168.20.60	k8s-master兼minion	kubernetes-v1.2.4、etcd-v2.3.2、flAnnel-v0.5.5、docker-v1.11.2	@H_867_197@etcd @H_867_197@flAnnel @H_867_197@docker @H_867_197@kube-apiserver @H_867_197@kube-controller-manager @H_867_197@kube-scheduler kubelet kube-proxy
192.168.20.61	k8s-minion1	kubernetes-v1.2.4、etcd-v2.3.2、flAnnel-v0.5.5、docker-v1.11.2	etcd flAnnel docker kubelet kube-proxy
192.168.20.62	k8s-minion2	kubernetes-v1.2.4、etcd-v2.3.2、flAnnel-v0.5.5、docker-v1.11.2	etcd flAnnel docker kubelet kube-proxy

二、环境准备：

系统环境: CentOS-7.2

#yum update

#关闭firewalld，安装iptables

systemctlstopfirewalld.service

systemctldisablefirewalld.service

yum-yinstalliptables-services

systemctlrestartiptables.service

systemctlenableiptables.service

#关闭selinux

sed-i"s/SELINUX=enforcing/SELINUX=disabled/g"/etc/selinux/config

setenforce0

#添加repo

#tee /etc/yum.repos.d/docker.repo <<-'EOF'

[dockerrepo]

name=Docker Repository

baseurl=https://yum.dockerproject.org/repo/main/centos/7/

enabled=1

gpgcheck=1

gpgkey=https://yum.dockerproject.org/gpg

EOF

#yuminstalldocker-ENGIne

#使用内部私有仓库

@H_867_197@#vi/usr/lib/systemd/system/docker.service

ExecStart=/usr/bin/dockerdaemon--insecure-registry=192.168.4.231:5000-Hfd://

#启动docker

systemctl start docker

三、安装etcd集群（为k8s提供存储功能和强一致性保证）

tarzxfetcd-v2.3.2-linux-amd64.tar.gz

cdetcd-v2.3.2-linux-amd64

cp etcd* /usr/local/bin/

#注册系统服务

#vi/usr/lib/systemd/system/etcd.service

Description=etcd

[service]

Environment=ETCD_NAME=@H_616_477@k8s-master#节点名称，唯一。minion节点就对应改为主机名就好

@H_867_197@Environment=ETCD_data_dir=/var/lib/etcd #存储数据路径，如果集群出现问题，可以删除这个目录重新配。

@H_867_197@Environment=ETCD_INITIAL_ADVERTISE_PEER_URLS=http://192.168.20.60:7001@H_867_197@ #监听地址，其他机器按照本机IP地址修改

@H_867_197@Environment=ETCD_LISTEN_PEER_URLS=http://192.168.20.60:7001@H_867_197@#监听地址，其他机器按照本机IP地址修改

@H_867_197@Environment=ETCD_LISTEN_CLIENT_URLS=http://192.168.20.60:4001,http://127.0.0.1:4001@H_867_197@ #对外监听地址，其他机器按照本机IP地址修改

@H_867_197@Environment=ETCD_ADVERTISE_CLIENT_URLS=http://192.168.20.60:4001@H_867_197@#对外监听地址，其他机器按照本机IP地址修改

@H_867_197@Environment=ETCD_INITIAL_CLUSTER_TOKEN=etcd-k8s-1 #集群名称，三台节点统一

@H_867_197@Environment=ETCD_INITIAL_CLUSTER=k8s-master=http://192.168.20.60:7001,k8s-minion1=http://192.168.20.61:7001,k8s-minion2=http://192.168.20.62:7001@H_867_197@ #集群监控

@H_867_197@Environment=ETCD_INITIAL_CLUSTER_STATE=new

@H_867_197@ExecStart=/usr/local/bin/etcd

@H_867_197@[Install]

@H_867_197@WantedBy=multi-user.target

@H_867_197@#启动服务

@H_867_197@systemctl start etcd

@H_867_197@#检查etcd集群是否正常工作

@H_867_197@[root@k8s-monion2@H_867_197@etcd]#etcdctlcluster-health

@H_867_197@member2d3a022000105975ishealthy:gothealthyresultfromhttp://192.168.20.61:4001

@H_867_197@member34a68a46747ee684ishealthy:gothealthyresultfromhttp://192.168.20.62:4001

@H_867_197@memberfe9e66405caec791ishealthy:gothealthyresultfromhttp://192.168.20.60:4001

@H_867_197@clusterishealthy#出现这个说明已经正常启动了。

@H_867_197@

@H_867_197@

@H_867_197@#然后设置一下打通的内网网段范围

@H_867_197@etcdctl set /coreos.com/network/config '{ "Network": "172.20.0.0/16" }'

@H_867_197@四、安装启动FlAnnel（打通容器间网络，可实现容器跨主机互连）

@H_867_197@tarzxfflAnnel-0.5.5-linux-amd64.tar.gz

@H_867_197@mvflAnnel-0.5.5/usr/local/flAnnel

@H_867_197@cd /usr/local/flAnnel

@H_867_197@#注册系统服务

@H_867_197@#vi/usr/lib/systemd/system/flAnneld.service

@H_867_197@[Unit]

@H_867_197@Description=flAnnel

@H_867_197@After=etcd.service

@H_867_197@After=docker.service

@H_867_197@[service]

@H_867_197@EnvironmentFile=/etc/sysconfig/flAnneld

@H_867_197@ExecStart=/usr/local/flAnnel/flAnneld\

@H_867_197@-etcd-endpoints=${FLAnnEL_ETCD}$FLAnnEL_OPTIONS

@H_867_197@[Install]

@H_867_197@WantedBy=multi-user.target

@H_867_197@#新建配置文件

@H_867_197@#vi /etc/sysconfig/flAnneld

@H_867_197@FLAnnEL_ETCD="http://192.168.20.60:4001,http://192.168.20.61:4001,http://192.168.20.62:4001@H_867_197@"

@H_867_197@

@H_867_197@#启动服务

@H_867_197@systemctlstartflAnneld

@H_867_197@mk-docker-opts.sh-i

source/run/flAnnel/subnet.env

@H_867_197@ifconfigdocker0${FLAnnEL_subnet}

@H_867_197@systemctlrestartdocker

@H_867_197@#验证是否成功

五、安装kubernets

1.下载源码包

cd /usr/local/

git clone

https://github.com/kubernetes/kubernetes.git

cd kubernetes/server/

tarzxf kubernetes-server-linux-amd64.tar.gz

cd kubernetes/server/bin

cpkube-apiserverkubectlkube-schedulerkube-controller-managerkube-proxy

kubelet

/usr/local/bin/

2.注册系统服务

#vi/usr/lib/systemd/system/kubelet.service

[Unit]

Description=KubernetesKubeletServer

Documentation=https://github.com/kubernetes/kubernetes

[service]

EnvironmentFile=/etc/kubernetes/config

EnvironmentFile=/etc/kubernetes/kubelet

User=root

ExecStart=/usr/local/bin/kubelet\

$KUBE_LOGTOSTDERR\

$KUBE_LOG_LEVEL\

$KUBE_ALLOW_PRIV\

$KUBELEt_address\

$KUBELET_PORT\

$KUBELET_HOSTNAME\

$KUBELET_API_SERVER\

$KUBELET_ARGS

Restart=on-failure

LimitNOFILE=65536

[Install]

WantedBy=multi-user.target

#vi/usr/lib/systemd/system/kube-proxy.service

[Unit]

Description=KubernetesKube-proxyServer

Documentation=https://github.com/kubernetes/kubernetes

[service]

EnvironmentFile=/etc/kubernetes/config

EnvironmentFile=/etc/kubernetes/kube-proxy

User=root

ExecStart=/usr/local/bin/kube-proxy\

$KUBE_LOGTOSTDERR\

$KUBE_LOG_LEVEL\

$KUBE_MASTER\

$KUBE_PROXY_ARGS

Restart=on-failure

LimitNOFILE=65536

[Install]

WantedBy=multi-user.target

3.创建配置文件

@H_32_14@mkdir /etc/kubernetes

vi /etc/kubernetes/config

#Commaseparatedlistofnodesintheetcdcluster

KUBE_ETCD_SERVERS="--etcd-servers=http://192.168.20.60:4001"

#loggingtostderrmeanswegetiTinthesystemdjournal

KUBE_LOGTOSTDERR="--logtostderr=true"

#journalmessagelevel,0isdebug

KUBE_LOG_LEVEL="--v=0"

#Shouldthisclusterbeallowedtorunprivilegeddockercontainers

#KUBE_ALLOW_PRIV="--allow-privileged=false"

KUBE_ALLOW_PRIV="--allow-privileged=true"

vi/etc/kubernetes/kubelet

#Theaddressfortheinfoservertoserveon

KUBELEt_address="--address=0.0.0.0"

#Theportfortheinfoservertoserveon

KUBELET_PORT="--port=10250"

#Youmayleavethisblanktousetheactualhostname

KUBELET_HOSTNAME="--hostname-override=192.168.20.60" #master，minion节点填本机的IP

#LOCATIOnoftheapi-server

KUBELET_API_SERVER="--api-servers=http://192.168.20.60:8080"

#Addyourown!

KUBELET_ARGS="--cluster-dns=192.168.20.64--cluster-domain=cluster.local" #后面使用dns插件会用到

vi/etc/kubernetes/kube-proxy

#HowtheReplicationcontrollerandschedulerfindthekube-apiserver

KUBE_MASTER="--master=http://192.168.20.60:8080"

#Addyourown!

KUBE_PROXY_ARGS="--proxy-mode=userspace" #代理模式，这里使用userspace。而iptables模式效率比较高，但要注意你的内核版本和iptables的版本是否符合要求，要不然会出错。

关于代理模式的选择，可以看国外友人的解释：

http://stackoverflow.com/questions/36088224/what-does-userspace-mode-means-in-kube-proxys-proxy-mode?rq=1

4.以上服务需要在所有节点上启动，下面的是master节点另外需要的服务：

kube-apiserver

、

kube-controller-manager

、

kube-scheduler

4.1、配置相关服务

#vi/usr/lib/systemd/system/kube-apiserver.service

After=etcd.service

[service]

EnvironmentFile=-/etc/kubernetes/config

EnvironmentFile=-/etc/kubernetes/apiserver

User=root

ExecStart=/usr/local/bin/kube-apiserver\

$KUBE_LOGTOSTDERR\

$KUBE_LOG_LEVEL\

$KUBE_ETCD_SERVERS\

$KUBE_API_ADDRESS\

$KUBE_API_PORT\

$KUBELET_PORT\

$KUBE_ALLOW_PRIV\

$KUBE_serviCE_ADDRESSES\

$KUBE_ADMISSION_CONTROL\

$KUBE_API_ARGS

Restart=on-failure

Type=notify

LimitNOFILE=65536

[Install]

WantedBy=multi-user.target

#vi/usr/lib/systemd/system/kube-controller-manager.service

[Unit]

Description=KubernetesControlleRMANager

Documentation=https://github.com/GoogleCloudPlatform/kubernetes

[service]

EnvironmentFile=-/etc/kubernetes/config

EnvironmentFile=-/etc/kubernetes/controller-manager

User=root

ExecStart=/usr/local/bin/kube-controller-manager\

$KUBE_LOGTOSTDERR\

$KUBE_LOG_LEVEL\

$KUBE_MASTER\

$KUBE_CONTROLLER_MANAGER_ARGS

Restart=on-failure

LimitNOFILE=65536

[Install]

WantedBy=multi-user.target

#vi/usr/lib/systemd/system/kube-scheduler.service

[Unit]

Description=KubernetesschedulerPlugin

Documentation=https://github.com/GoogleCloudPlatform/kubernetes

[service]

EnvironmentFile=-/etc/kubernetes/config

EnvironmentFile=-/etc/kubernetes/scheduler

User=root

ExecStart=/usr/local/bin/kube-scheduler\

$KUBE_LOGTOSTDERR\

$KUBE_LOG_LEVEL\

$KUBE_MASTER\

$KUBE_scheDULER_ARGS

Restart=on-failure

LimitNOFILE=65536

[Install]

@H_867_197@4.2、配置文件

vi /etc/kubernetes/apiserver

#TheaddressontHelocalservertolistento.

KUBE_API_ADDRESS="--address=0.0.0.0"

#TheportontHelocalservertolistenon.

KUBE_API_PORT="--port=8080"

#HowtheReplicationcontrollerandschedulerfindthekube-apiserver

KUBE_MASTER="--master=http://192.168.20.60:8080"

#Portkubeletslistenon

KUBELET_PORT="--kubelet-port=10250"

#Addressrangetouseforservices

KUBE_serviCE_ADDRESSES="--service-cluster-ip-range=192.168.20.0/24"

#KUBE_serviCE_ADDRESSES="--service-cluster-ip-range=10.254.0.0/16"

#Addyourown!

KUBE_API_ARGS=""

vi /etc/kubernetes/controller-manager

#HowtheReplicationcontrollerandschedulerfindthekube-apiserver

KUBE_MASTER="--master=http://192.168.20.60:8080"

#Addyourown!

KUBE_CONTROLLER_MANAGER_ARGS=""

vi /etc/kubernetes/scheduler

#HowtheReplicationcontrollerandschedulerfindthekube-apiserver

KUBE_MASTER="--master=http://192.168.20.60:8080"

#Addyourown!

KUBE_scheDULER_ARGS=""

更多配置项可以参考官方文档：

http://kubernetes.io/docs/admin/kube-proxy/

#启动master服务

systemctlstartkubelet

systemctlstartkube-proxy

systemctlstartkube-apiserver

systemctlstartkube-controller-manager

systemctlstartkube-scheduler

#启动minion服务

systemctlstartkubelet

systemctlstartkube-proxy

#检查服务是否启动正常

[root@k8s-masterbin]#kubectlgetno

NAMESTATUSAGE

192.168.20.60Ready24s

192.168.20.61Ready46s

192.168.20.62Ready35s

#重启命令

systemctlrestartkubelet

systemctlrestartkube-proxy

systemctlrestartkube-apiserver

systemctlrestart kube-controller-manager

systemctlrestartkube-scheduler

#填坑

pause gcr.io 被墙，没有这个镜像k8s应用不了，报下面错误：

image pull Failed for gcr.io/google_containers/pause:2.0

使用docker hub的镜像代替，或者下到本地仓库，然后再重新打tag，并且每个节点都需要这个镜像

docker pullkubernetes/pause

dockertag kubernetes/pause gcr.io/google_containers/pause:2.0

[root@k8s-masteraddons]#dockerimages

REPOSITORYTAGIMAGEIDCREATEDSIZE

192.168.4.231:5000/pause2.02b58359142b09monthsago350.2kB

gcr.io/google_containers/pause2.02b58359142b09monthsago350.2kB

5.官方源码包里有一些插件，如：监控面板、dns

cd /usr/local/kubernetes/cluster/addons/

5.1、dashboard 监控面板插件

cd /usr/local/kubernetes/cluster/addons/dashboard

下面有两个文件：

=============================================================

dashboard-controller.yaml #用来设置部署应用，如：副本数，使用镜像，资源控制等等

apiVersion:v1

kind:ReplicationController

Metadata:

#Keepthenameinsyncwithimageversionand

#gce/coreos/kube-manifests/addons/dashboardcounterparts

name:kubernetes-dashboard-v1.0.1

namespace:kube-system

labels:

k8s-app:kubernetes-dashboard

version:v1.0.1

kubernetes.io/cluster-service:"true"

spec:

replicas:1 #副本数量

SELEctor:

k8s-app:kubernetes-dashboard

template:

Metadata:

labels:

k8s-app:kubernetes-dashboard

version:v1.0.1

kubernetes.io/cluster-service:"true"

spec:

containers:

-name:kubernetes-dashboard

image:192.168.4.231:5000/kubernetes-dashboard:v1.0.1

resources:

#keeprequest=limittokeepthiscontaineringuaranteedclass

limits:

cpu:100m

@H_32_14@memory:50Mi

requests:

cpu:100m

@H_32_14@memory:50Mi

ports:

-containerPort:9090

args:

---apiserver-host=http://192.168.20.60:8080#这里需要注意，不加这个参数，会默认去找localhost，而不是去master那里取。还有就是这个配置文件各项缩减问题，空格。

livenessProbe:

httpGet:

path:/

port:9090

initialDelaySeconds:30

timeoutSeconds:30

========================================================

dashboard-service.yaml #提供外部访问服务

apiVersion:v1

kind:service

Metadata:

name:kubernetes-dashboard

namespace:kube-system

labels:

k8s-app:kubernetes-dashboard

kubernetes.io/cluster-service:"true"

spec:

SELEctor:

k8s-app:kubernetes-dashboard

ports:

-port:80

targetPort:9090

=========================================================

kubectlcreate-f./ #创建服务

kubectl--namespace=kube-systemgetpo #查看系统服务启动状态

s.width=650;" src="http://img.code.cc/vcimg/static/loading.png" style="BACkground:url("/e/u261/lang/zh-cn/images/localimage.png") no-repeat center;border:1px solid #ddd;" src="http://chenguomin.blog.51cto.come/u261/themes/default/images/spacer.gif">

kubectl--namespace=kube-systemgetpo-owide #查看系统服务起在哪个节点

s.width=650;" src="http://img.code.cc/vcimg/static/loading.png" style="BACkground:url("/e/u261/lang/zh-cn/images/localimage.png") no-repeat center;border:1px solid #ddd;" src="http://chenguomin.blog.51cto.come/u261/themes/default/images/spacer.gif">

#若想删除，可以执行下面命令

kubectldelete-f./

在浏览器输入： http://192.168.20.60:8080/ui/

@H_618_1727@5.2、DNS 插件安装

@H_419_1722@#使用ip地址方式不太容易记忆，集群内可以使用dns绑定ip并自动更新维护。

cd/usr/local/kubernetes/cluster/addons/dns

cpskydns-rc.yaml.in /opt/dns/skydns-rc.yaml

cpskydns-svc.yaml.in /opt/dns/skydns-svc.yaml

@H_807_1772@#/opt/dns/skydns-rc.yaml 文件

apiVersion:v1

kind:ReplicationController

Metadata:

name:kube-dns-v11

namespace:kube-system

labels:

k8s-app:kube-dns

version:v11

kubernetes.io/cluster-service:"true"

spec:

replicas:1

SELEctor:

k8s-app:kube-dns

version:v11

template:

Metadata:

labels:

k8s-app:kube-dns

version:v11

kubernetes.io/cluster-service:"true"

spec:

containers:

-name:etcd

image:192.168.4.231:5000/etcd-amd64:2.2.1 #我都是先把官方镜像下载到本地仓库

resources:

#TODO:Setmemorylimitswhenwe'veprofiledthecontainerforlarge

#clusters,thensetrequest=limittokeepthiscontainerin

#guaranteedclass.Currently,thiscontainerfallsintothe

#"burstable"categorysothekubeletdoesn'tBACkofffromrestarTingit.

limits:

cpu:100m

@H_201_1737@memory:500Mi

requests:

cpu:100m

@H_201_1737@memory:50Mi

command:

-/usr/local/bin/etcd

--data-dir

-/var/etcd/data

--listen-client-urls

-http://127.0.0.1:2379,http://127.0.0.1:4001

--advertise-client-urls

-http://127.0.0.1:2379,http://127.0.0.1:4001

--initial-cluster-token

-skydns-etcd

volumeMounts:

-name:etcd-storage

@H_201_1737@mountPath:/var/etcd/data

-name:kube2sky

image:192.168.4.231:5000/kube2sky:1.14 #本地仓库取

resources:

#TODO:Setmemorylimitswhenwe'veprofiledthecontainerforlarge

#clusters,thiscontainerfallsintothe

#"burstable"categorysothekubeletdoesn'tBACkofffromrestarTingit.

limits:

cpu:100m

#Kube2skywatchesallpods.

@H_201_1737@memory:200Mi

requests:

cpu:100m

@H_201_1737@memory:50Mi

livenessProbe:

httpGet:

path:/healthz

port:8080

scheR_501_11845@e:http

initialDelaySeconds:60

timeoutSeconds:5

successThreshold:1

failureThreshold:5

readinessProbe:

httpGet:

path:/readiness

port:8081

scheR_501_11845@e:http

#wepollonpodstartupfortheKubernetesmasterserviceand

#onlysetupthe/readinesshttpserveronCEThat'savailable.

initialDelaySeconds:30

timeoutSeconds:5

args:

#command="/kube2sky"

---domain=cluster.local #一个坑，要和/etc/kubernetes/kubelet 内的一致

---kube_master_url=http://192.168.20.60:8080 #master管理节点

-name:skydns

image:192.168.4.231:5000/skydns:2015-10-13-8c72f8c

resources:

#TODO:Setmemorylimitswhenwe'veprofiledthecontainerforlarge

#clusters,thiscontainerfallsintothe

#"burstable"categorysothekubeletdoesn'tBACkofffromrestarTingit.

limits:

cpu:100m

@H_201_1737@memory:200Mi

requests:

cpu:100m

@H_201_1737@memory:50Mi

args:

#command="/skydns"

--machines=http://127.0.0.1:4001

--addr=0.0.0.0:53

--ns-rotate=false

--domain=cluster.local. #另一个坑！！ 后面要带"."

ports:

-containerPort:53

name:dns

protocol:UDP

-containerPort:53

name:dns-tcp

protocol:TCP

-name:healthz

image:192.168.4.231:5000/exechealthz:1.0 #本地仓库取镜像

resources:

#keeprequest=limittokeepthiscontaineringuaranteedclass

limits:

cpu:10m

@H_201_1737@memory:20Mi

requests:

cpu:10m

@H_201_1737@memory:20Mi

args:

--cmd=nslookupkubernetes.default.svc.cluster.local127.0.0.1>/dev/null #还是这个坑

--port=8080

ports:

-containerPort:8080

protocol:TCP

volumes:

-name:etcd-storage

emptyDir:{}

dnsPolicy:Default#Don'tuseclusterDNs.

========================================================================

#/opt/dns/skydns-svc.yaml 文件

apiVersion:v1

kind:service

Metadata:

name:kube-dns

namespace:kube-system

labels:

k8s-app:kube-dns

kubernetes.io/cluster-service:"true"

kubernetes.io/name:"KubeDNS"

spec:

SELEctor:

k8s-app:kube-dns

clusterIP:192.168.20.100

ports:

-name:dns

port:53

protocol:UDP

-name:dns-tcp

port:53

protocol:TCP

==================================================================

#启动

cd /opt/dns/

kubectl create -f ./

#查看pod启动状态，看到4/4 个服务都启动完成，那就可以进行下一步验证阶段。

kubectl--namespace=kube-systemgetpod-owide

@H_618_1727@转官网的验证方法：

网址：https://github.com/kubernetes/kubernetes/blob/release-1.2/cluster/addons/dns/README.md#userconsent#

How do I test if it is working?

First deploy DNS as described above.

1 Create a simple Pod to use as a test environment.

Create a file named busyBox.yaml with the following contents:

apiVersion:v1kind:PodMetadata:name:busyBoxnamespace:defaultspec:containers:-image:busyBoxcommand:-sleep-"3600"imagePullPolicy:IfNotPresentname:busyBoxrestartPolicy:Always

Then create a pod using this file:

kubectlcreate-fbusyBox.yaml

2 Wait for this pod to go into the running state.

You can get its status with:

@H_510_2489@kubectlgetpodsbusyBox

You should see:

@H_510_2489@NAMEREADYSTATUSRESTARTSAGE busyBox1/1Running0<some-time>

3 Validate DNS works

Once that pod is running,you can exec nslookup in that environment:

@H_510_2489@kubectlexecbusyBox--nslookupkubernetes.default

You should see something like:

@H_510_2489@Server:10.0.0.10 Address1:10.0.0.10 Name:kubernetes.default Address1:10.0.0.1

If you see that,DNS is working correctly.

5.3、manage插件

参考：http://my.oschina.net/fufangchun/blog/703985

@H_791_2511@mkdir /opt/k8s-manage

cd /opt/k8s-manage

================================================

#catk8s-manager-rc.yaml

apiVersion:v1

kind:ReplicationController

Metadata:

name:k8s-manager

namespace:kube-system

labels:

app:k8s-manager

spec:

replicas:1

SELEctor:

app:k8s-manager

template:

Metadata:

labels:

app:k8s-manager

spec:

containers:

-image:@H_361_1540@mlamina/k8s-manager:latest

name:k8s-manager

resources:

limits:

cpu:100m

@H_32_14@memory:50Mi

ports:

-containerPort:80

name:http

=================================================

#catk8s-manager-svr.yaml

apiVersion:v1

kind:service

Metadata:

name:k8s-manager

namespace:kube-system

labels:

app:k8s-manager

spec:

ports:

-port:80

targetPort:http

SELEctor:

app:k8s-manager

=================================================

#启动

kubectl create -f ./

浏览器访问：

http://192.168.20.60:8080/api/v1/proxy/namespaces/kube-system/services/k8s-manager

@L_616_197@

实例演示

1.搭zookeeper、activeMQ、redis、mongodb服务

@H_457_2689@mkdir /opt/service/

cd /opt/service

==========================================================

#cat service.yaml

apiVersion:v1

kind:service

Metadata:

name:zk-amq-rds-mgd #服务名称

labels:

run:zk-amq-rds-mgd

spec:

type:NodePort

ports:

-port:2181 #标识

nodePort:31656#master节点对外服务端口

targetPort:2181 #容器内部端口

protocol:TCP #协议类型

name:zk-app #表示名

-port:8161

nodePort:31654

targetPort:8161

protocol:TCP

name:amq-http

-port:61616

@H_502_2774@nodePort:31655

@H_502_2774@targetPort:61616

@H_502_2774@protocol:TCP

@H_502_2774@name:amq-app

-port:27017

@H_491_2797@nodePort:31653

@H_491_2797@targetPort:27017

@H_491_2797@protocol:TCP

@H_491_2797@name:mgd-app

-port:6379

nodePort:31652

targetPort:6379

protocol:TCP

name:rds-app

SELEctor:

run:zk-amq-rds-mgd

---

#apiVersion:extensions/v1beta1

apiVersion:v1

kind:ReplicationController

Metadata:

name:zk-amq-rds-mgd

spec:

replicas:2 #两个副本

template:

Metadata:

labels:

run:zk-amq-rds-mgd

spec:

containers:

-name:zookeeper #应用名称

image:192.168.4.231:5000/zookeeper:0524 #使用本地镜像

imagePullPolicy:IfNotPresent #自动分配到性能好的节点，去拉镜像并启动容器。

ports:

-containerPort:2181 #容器内部服务端口

env:

-name:LANG

value:en_US.UTF-8

volumeMounts:

-mountPath:/tmp/zookeeper #容器内部挂载点

name:zookeeper-d#挂载名称，要与下面配置外部挂载点一致

-name:activemq

image:192.168.4.231:5000/activemq:v2

imagePullPolicy:IfNotPresent

ports:

-containerPort:8161

-containerPort:61616

volumeMounts:

-mountPath:/opt/apache-activemq-5.10.2/data

name:activemq-d

@H_491_2797@-name:mongodb

@H_491_2797@image:192.168.4.231:5000/mongodb:3.0.6

@H_491_2797@imagePullPolicy:IfNotPresent

@H_491_2797@ports:

@H_491_2797@-containerPort:27017

@H_491_2797@volumeMounts:

@H_491_2797@-mountPath:/var/lib/mongo

@H_491_2797@name:mongodb-d

-name:redis

image:192.168.4.231:5000/redis:2.8.25

imagePullPolicy:IfNotPresent

ports:

-containerPort:6379

volumeMounts:

-mountPath:/opt/redis/var

name:redis-d

volumes:

-hostPath:

path:/mnt/mfs/service/zookeeper/data #宿主机挂载点，我这里用了分布式共享存储(MooseFS)，这样可以保证多个副本数据的一致性。

name:zookeeper-d

-hostPath:

path:/mnt/mfs/service/activemq/data

-hostPath:

@H_491_2797@path:/mnt/mfs/service/mongodb/data

-hostPath:

path:/mnt/mfs/service/redis/data

name:redis-d

===========================================================================================

#创建服务

kubectl create -f ./

参考文献：http://my.oschina.net/jayqqaa12/blog/693919#userconsent#