一、环境准备
1、准备三台虚拟机,具体信息如下,配置好root账户,安装好docker,安装方法参见https://www.cnblogs.com/liangyuntao-ts/p/10657009.htm
系统类型 IP地址 节点角色 CPU Memory Hostname
centos7 192.168.100.101 worker 1 2G work01
centos7 192.168.100.102 master 1 2G master
centos7 192.168.100.103 worker 1 2G work02
2、三台服务器启动docker
[root@server02 ~]# systemctl start docker
[root@server02 ~]# systemctl enable docker
[root@server02 ~]# docker version
Client:
Version: 18.09.6
API version: 1.39
Go version: go1.10.8
Git commit: 481bc77156
Built: Sat May 4 02:34:58 2019
OS/Arch: linux/amd64
Experimental: false
Server: Docker Engine - Community
Engine:
Version: 18.09.6
API version: 1.39 (minimum version 1.12)
Go version: go1.10.8
Git commit: 481bc77
Built: Sat May 4 02:02:43 2019
OS/Arch: linux/amd64
Experimental: false
3、系统设置,关闭防火墙,selinux,设置路由转发,不对bridge数据进行处理
[root@server02 ~]# systemctl stop firewalld
[root@server02 ~]# systemctl disable firewalld
Removed symlink /etc/systemd/system/multi-user.target.wants/firewalld.service.
Removed symlink /etc/systemd/system/dbus-org.fedoraproject.FirewallD1.service.
写入/生效配置文件
[root@server02 ~]# cat <<EOF > /etc/sysctl.d/k8s.conf
net.ipv4.ip_forward = 1
net.bridge.bridge-nf-call-ip6tables = 1
net.bridge.bridge-nf-call-iptables = 1
EOF
[root@server02 ~]# sysctl -p /etc/sysctl.d/k8s.conf
4、配置host文件
配置host,使每个Node都可以通过名字解析到ip地址
[root@server02 ~]# vi /etc/hosts
#加入如下片段(ip地址和servername替换成自己的)
192.168.100.101 server01
192.168.100.102 master
192.168.100.103 server02
5、准备二进制文件
下载地址:
v1.9 链接:https://pan.baidu.com/s/13izNNZ3Bkem61Zemhkj8gQ
提取码:0ykv
注:下载完成后,将二进制命令文件上传至所有的服务器的/home目录下,v1.9.8的二进制文件在kubernetes/server/bin下
6、准备配置文件
到home目录下载项目
[root@server02 ~]# cd /home
[root@server02 ~]# git clone https://github.com/liuyi01/kubernetes-starter.git
#看看git内容
[root@server02 ~]# cd ~/kubernetes-starter && ls
apps config.properties docs gen-config.sh kubernetes-simple kubernetes-with-ca README.md service-config
7、修改配置文件,生产适应环境的配置,三台服务器都需要设置
[root@server02 home]# vim kubernetes-starter/config.properties
#kubernetes二进制文件目录,eg: /home/michael/bin
BIN_PATH=/home/bin
#当前节点ip, eg: 192.168.1.102
NODE_IP=192.168.100.102
#etcd服务集群列表, eg: http://192.168.1.102:2379
#如果已有etcd集群可以填写现有的。没有的话填写:http://${MASTER_IP}:2379 (MASTER_IP自行替换成自己的主节点ip)
##如果用了证书,就要填写https://${MASTER_IP}:2379 (MASTER_IP自行替换成自己的主节点ip)
ETCD_ENDPOINTS=http://192.168.100.102:2379
#kubernetes主节点ip地址, eg: 192.168.1.102
MASTER_IP=192.168.100.102
#config.properties配置文件根据自己的配置进行设置
[root@server02 home]# ./gen-config.sh simple
[root@server02 home]# mv kubernetes-bins/ bin 将该解压的文件夹修改为/home/bin ,在将该路径加入环境遍历中#v1.9.8和1.9命令文件位置不同,按1.9.8情况修改,懒得再整理
[root@server02 home]# vi ~/.bash_profile
PATH=$PATH:$HOME/bin
[root@server02 home]# export PATH=$PATH:/home/bin //设置环境变量
二、基础服务部署
1、ETCD服务部署,二进制文件已经准备好,现在把它做成系统服务并启动(主节点操作)
把服务配置文件copy到系统服务目录
[root@server02 ~]# cp /home/kubernetes-starter/target/master-node/etcd.service /lib/systemd/system/
#enable服务
[root@server02 ~]# systemctl enable etcd.service
#创建工作目录(保存数据的地方)
[root@server02 ~]# mkdir -p /var/lib/etcd
# 启动服务
[root@server02 ~]# systemctl start etcd
# 查看服务日志,看是否有错误信息,确保服务正常
[root@server02 ~]# journalctl -f -u etcd.service
5月 18 12:17:31 server02 etcd[2179]: dialing to target with scheme: ""
5月 18 12:17:31 server02 etcd[2179]: could not get resolver for scheme: ""
5月 18 12:17:31 server02 etcd[2179]: serving insecure client requests on 192.168.100.102:2379, this is strongly discouraged!
5月 18 12:17:31 server02 etcd[2179]: ready to serve client requests
5月 18 12:17:31 server02 etcd[2179]: dialing to target with scheme: ""
5月 18 12:17:31 server02 etcd[2179]: could not get resolver for scheme: ""
5月 18 12:17:31 server02 etcd[2179]: serving insecure client requests on 127.0.0.1:2379, this is strongly discouraged!
5月 18 12:17:31 server02 etcd[2179]: set the initial cluster version to 3.2
5月 18 12:17:31 server02 etcd[2179]: enabled capabilities for version 3.2
5月 18 12:17:31 server02 systemd[1]: Started Etcd Server.
####ETCD服务正常启动
2、部署APIServer(主节点)
简介:
kube-apiserver是Kubernetes最重要的核心组件之一,主要提供以下的功能
- 提供集群管理的REST API接口,包括认证授权(我们现在没有用到)数据校验以及集群状态变更等
- 提供其他模块之间的数据交互和通信的枢纽(其他模块通过API Server查询或修改数据,只有API Server才直接操作etcd
[root@server02 ~]# cd /home/
[root@server02 home]# cp kubernetes-starter/target/master-node/kube-apiserver.service /lib/systemd/system/
[root@server02 home]# systemctl enable kube-apiserver.service
Created symlink from /etc/systemd/system/multi-user.target.wants/kube-apiserver.service to /usr/lib/systemd/system/kube-apiserver.service.
[root@server02 home]# systemctl start kube-apiserver
[root@server02 home]# journalctl -f -u kube-apiserver
-- Logs begin at 六 2019-05-18 11:47:54 CST. --
5月 18 12:57:19 server02 kube-apiserver[2333]: I0518 12:57:19.688480 2333 wrap.go:42] PUT /apis/apiregistration.k8s.io/v1beta1/apiservices/v1beta1.certificates.k8s.io/status: (46.900994ms) 200 [[kube-apiserver/v1.9.0 (linux/amd64) kubernetes/925c127] 127.0.0.1:49330]
5月 18 12:57:19 server02 kube-apiserver[2333]: I0518 12:57:19.691365 2333 wrap.go:42] PUT /apis/apiregistration.k8s.io/v1beta1/apiservices/v1beta1.policy/status: (40.847972ms) 200 [[kube-apiserver/v1.9.0 (linux/amd64) kubernetes/925c127] 127.0.0.1:49330]
5月 18 12:57:19 server02 kube-apiserver[2333]: I0518 12:57:19.692039 2333 wrap.go:42] PUT /apis/apiregistration.k8s.io/v1beta1/apiservices/v1.storage.k8s.io/status: (41.81334ms) 200 [[kube-apiserver/v1.9.0 (linux/amd64) kubernetes/925c127] 127.0.0.1:49330]
5月 18 12:57:19 server02 kube-apiserver[2333]: I0518 12:57:19.703752 2333 wrap.go:42] PUT /apis/apiregistration.k8s.io/v1beta1/apiservices/v1.rbac.authorization.k8s.io/status: (11.64213ms) 200 [[kube-apiserver/v1.9.0 (linux/amd64) kubernetes/925c127] 127.0.0.1:49330]
5月 18 12:57:19 server02 kube-apiserver[2333]: I0518 12:57:19.704980 2333 wrap.go:42] PUT /apis/apiregistration.k8s.io/v1beta1/apiservices/v1.networking.k8s.io/status: (13.967816ms) 200 [[kube-apiserver/v1.9.0 (linux/amd64) kubernetes/925c127] 127.0.0.1:49330]
5月 18 12:57:19 server02 kube-apiserver[2333]: I0518 12:57:19.710226 2333 wrap.go:42] PUT /apis/apiregistration.k8s.io/v1beta1/apiservices/v1beta1.rbac.authorization.k8s.io/status: (5.19179ms) 200 [[kube-apiserver/v1.9.0 (linux/amd64) kubernetes/925c127] 127.0.0.1:49330]
5月 18 12:57:19 server02 kube-apiserver[2333]: I0518 12:57:19.710252 2333 wrap.go:42] PUT /apis/apiregistration.k8s.io/v1beta1/apiservices/v1beta1.storage.k8s.io/status: (5.695826ms) 200 [[kube-apiserver/v1.9.0 (linux/amd64) kubernetes/925c127] 127.0.0.1:49330]
5月 18 12:57:29 server02 kube-apiserver[2333]: I0518 12:57:29.559583 2333 wrap.go:42] GET /api/v1/namespaces/default: (4.524421ms) 200 [[kube-apiserver/v1.9.0 (linux/amd64) kubernetes/925c127] 127.0.0.1:49330]
5月 18 12:57:29 server02 kube-apiserver[2333]: I0518 12:57:29.563896 2333 wrap.go:42] GET /api/v1/namespaces/default/services/kubernetes: (2.544183ms) 200 [[kube-apiserver/v1.9.0 (linux/amd64) kubernetes/925c127] 127.0.0.1:49330]
5月 18 12:57:29 server02 kube-apiserver[2333]: I0518 12:57:29.566296 2333 wrap.go:42] GET /api/v1/namespaces/default/endpoints/kubernetes: (1.280719ms) 200 [[kube-apiserver/v1.9.0 (linux/amd64) kubernetes/925c127] 127.0.0.1:49330]
####日志都是提示的,没有异常
查看端口是否起来
[root@server02 home]# netstat -ntlp
Active Internet connections (only servers)
Proto Recv-Q Send-Q Local Address Foreign Address State PID/Program name
tcp 0 0 192.168.100.102:2379 0.0.0.0:* LISTEN 2179/etcd
tcp 0 0 127.0.0.1:2379 0.0.0.0:* LISTEN 2179/etcd
tcp 0 0 127.0.0.1:2380 0.0.0.0:* LISTEN 2179/etcd
tcp 0 0 0.0.0.0:22 0.0.0.0:* LISTEN 836/sshd
tcp 0 0 127.0.0.1:25 0.0.0.0:* LISTEN 1069/master
tcp6 0 0 :::6443 :::* LISTEN 2333/kube-apiserver
tcp6 0 0 :::8080 :::* LISTEN 2333/kube-apiserver
tcp6 0 0 :::22 :::* LISTEN 836/sshd
tcp6 0 0 ::1:25 :::* LISTEN 1069/master
3、部署ControllerManager (主节点)
Controller Manager由kube-controller-manager和cloud-controller-manager组成,是Kubernetes的大脑,它通过apiserver监控整个集群的状态,并确保集群处于预期的工作状态。 kube-controller-manager由一系列的控制器组成,像Replication Controller控制副本,Node Controller节点控制,Deployment Controller管理deployment等等 cloud-controller-manager在Kubernetes启用Cloud Provider的时候才需要,用来配合云服务提供商的控制
[root@server02 home]# cp kubernetes-starter/target/master-node/kube-controller-manager.service /lib/systemd/system/
[root@server02 home]# systemctl enable kube-controller-manager.service
Created symlink from /etc/systemd/system/multi-user.target.wants/kube-controller-manager.service to /usr/lib/systemd/system/kube-controller-manager.service.
[root@server02 home]# systemctl start kube-controller-manager.service
[root@server02 home]# journalctl -f -u kube-controller-manager
-- Logs begin at 六 2019-05-18 11:47:54 CST. --
5月 18 13:26:11 server02 systemd[1]: Started Kubernetes Controller Manager.
5月 18 13:26:11 server02 systemd[1]: Starting Kubernetes Controller Manager...
4、部署Scheduler(主节点)
kube-scheduler负责分配调度Pod到集群内的节点上,它监听kube-apiserver,查询还未分配Node的Pod,然后根据调度策略为这些Pod分配节点。我们前面讲到的kubernetes的各种调度策略就是它实现的。
[root@server02 home]# cp kubernetes-starter/target/master-node/kube-scheduler.service /lib/systemd/system/
[root@server02 home]# systemctl enable kube-schedule.service
Failed to execute operation: No such file or directory
[root@server02 home]# systemctl enable kube-scheduler.service
Created symlink from /etc/systemd/system/multi-user.target.wants/kube-scheduler.service to /usr/lib/systemd/system/kube-scheduler.service.
[root@server02 home]# systemctl start kube-scheduler.service
[root@server02 home]# journalctl -f -u kube-scheduler
-- Logs begin at 六 2019-05-18 11:47:54 CST. --
5月 18 13:29:09 server02 systemd[1]: Starting Kubernetes Scheduler...
5月 18 13:29:10 server02 kube-scheduler[2430]: W0518 13:29:10.675474 2430 server.go:159] WARNING: all flags than --config are deprecated. Please begin using a config file ASAP.
5月 18 13:29:10 server02 kube-scheduler[2430]: I0518 13:29:10.728026 2430 server.go:551] Version: v1.9.0
5月 18 13:29:10 server02 kube-scheduler[2430]: I0518 13:29:10.729972 2430 factory.go:837] Creating scheduler from algorithm provider 'DefaultProvider'
5月 18 13:29:10 server02 kube-scheduler[2430]: I0518 13:29:10.730027 2430 factory.go:898] Creating scheduler with fit predicates 'map[MaxAzureDiskVolumeCount:{} NoDiskConflict:{} CheckNodeMemoryPressure:{} NoVolumeZoneConflict:{} MaxEBSVolumeCount:{} MaxGCEPDVolumeCount:{} MatchInterPodAffinity:{} GeneralPredicates:{} CheckNodeDiskPressure:{} CheckNodeCondition:{} PodToleratesNodeTaints:{} CheckVolumeBinding:{}]' and priority functions 'map[SelectorSpreadPriority:{} InterPodAffinityPriority:{} LeastRequestedPriority:{} BalancedResourceAllocation:{} NodePreferAvoidPodsPriority:{} NodeAffinityPriority:{} TaintTolerationPriority:{}]'
5月 18 13:29:10 server02 kube-scheduler[2430]: I0518 13:29:10.730923 2430 server.go:570] starting healthz server on 127.0.0.1:10251
5月 18 13:29:11 server02 kube-scheduler[2430]: I0518 13:29:11.642607 2430 controller_utils.go:1019] Waiting for caches to sync for scheduler controller
5月 18 13:29:11 server02 kube-scheduler[2430]: I0518 13:29:11.743117 2430 controller_utils.go:1026] Caches are synced for scheduler controller
5月 18 13:29:11 server02 kube-scheduler[2430]: I0518 13:29:11.766782 2430 leaderelection.go:174] attempting to acquire leader lease...
5月 18 13:29:11 server02 kube-scheduler[2430]: I0518 13:29:11.786299 2430 leaderelection.go:184] successfully acquired lease kube-system/kube-scheduler
5、下面配置kubectl命令(任意节点,这个我们也部署在主节点)
kubectl是Kubernetes的命令行工具,是Kubernetes用户和管理员必备的管理工具。 kubectl提供了大量的子命令,方便管理Kubernetes集群中的各种功能。
设置api-server和上下文
#指定apiserver地址(ip替换为你自己的api-server地址)
[root@server02 home]# kubectl config set-cluster kubernetes --server=http://192.168.100.102:8080
Cluster "kubernetes" set.
#指定设置上下文,指定cluster
[root@server02 home]# kubectl config set-context kubernetes --cluster=kubernetes
Context "kubernetes" created.
#选择默认的上下文
[root@server02 home]# kubectl config use-context kubernetes
Switched to context "kubernetes".
6、配置Kuberlet(work节点)
每个工作节点上都运行一个kubelet服务进程,默认监听10250端口,接收并执行master发来的指令,管理Pod及Pod中的容器。每个kubelet进程会在API Server上注册节点自身信息,定期向master节点汇报节点的资源使用情况,并通过cAdvisor监控节点和容器的资源。
#确保相关目录存在
[root@server01 home]# mkdir -p /var/lib/kubelet
[root@server01 home]# mkdir -p /etc/kubernetes
[root@server01 home]# mkdir -p /etc/cni/net.d
#复制kubelet服务配置文件
[root@server01 home]# cp kubernetes-starter/target/worker-node/kubelet.service /lib/systemd/system
#复制kubelet依赖的配置文件
[root@server01 home]# cp kubernetes-starter/target/worker-node/kubelet.kubeconfig /etc/kubernetes/
#复制kubelet用到的cni插件配置文件
[root@server01 home]# cp kubernetes-starter/target/worker-node/10-calico.conf /etc/cni/net.d/
[root@server01 home]# systemctl enable kubelet.service
Created symlink from /etc/systemd/system/multi-user.target.wants/kubelet.service to /usr/lib/systemd/system/kubelet.service.
[root@server01 home]# systemctl start kubelet
[root@server01 kubernetes-starter]# journalctl -f -u kubelet
-- Logs begin at 六 2019-05-18 11:47:58 CST. --
5月 18 14:17:59 server01 kubelet[8077]: I0518 14:17:59.886468 8077 manager.go:1178] Started watching for new ooms in manager
5月 18 14:17:59 server01 kubelet[8077]: I0518 14:17:59.939249 8077 kubelet_node_status.go:273] Setting node annotation to enable volume controller attach/detach
5月 18 14:17:59 server01 kubelet[8077]: I0518 14:17:59.976545 8077 kubelet_node_status.go:431] Recording NodeHasSufficientDisk event message for node 192.168.100.101
5月 18 14:17:59 server01 kubelet[8077]: I0518 14:17:59.976684 8077 kubelet_node_status.go:431] Recording NodeHasSufficientMemory event message for node 192.168.100.101
5月 18 14:17:59 server01 kubelet[8077]: I0518 14:17:59.976713 8077 kubelet_node_status.go:431] Recording NodeHasNoDiskPressure event message for node 192.168.100.101
5月 18 14:17:59 server01 kubelet[8077]: I0518 14:17:59.976736 8077 kubelet_node_status.go:82] Attempting to register node 192.168.100.101
5月 18 14:18:00 server01 kubelet[8077]: I0518 14:18:00.385257 8077 manager.go:329] Starting recovery of all containers
5月 18 14:18:00 server01 kubelet[8077]: I0518 14:18:00.773129 8077 kubelet_node_status.go:85] Successfully registered node 192.168.100.101
7、给集群增加Service的功能----Kube-Proxy(工作节点)
#确保工作目录存在
[root@server03 kubernetes-starter]# mkdir -p /var/lib/kube-proxy
#复制kube-proxy服务配置文件
[root@server03 kubernetes-starter]# cp target/worker-node/kube-proxy.service /lib/systemd/system/
#复制kube-proxy依赖的配置文件
[root@server03 kubernetes-starter]# cp target/worker-node/kube-proxy.kubeconfig /etc/kubernetes/
[root@server03 kubernetes-starter]# systemctl enable kube-proxy.service
[root@server03 kubernetes-starter]# systemctl start kube-proxy
[root@server03 kubernetes-starter]# journalctl -f -u kube-proxy
-- Logs begin at 四 2019-05-16 13:20:42 CST. --
5月 19 12:15:56 localhost.localdomain kube-proxy[28949]: I0519 12:15:56.425552 28949 conntrack.go:83] Setting conntrack hashsize to 32768
5月 19 12:15:56 localhost.localdomain kube-proxy[28949]: I0519 12:15:56.426442 28949 conntrack.go:98] Set sysctl 'net/netfilter/nf_conntrack_tcp_timeout_established' to 86400
5月 19 12:15:56 localhost.localdomain kube-proxy[28949]: I0519 12:15:56.426524 28949 conntrack.go:98] Set sysctl 'net/netfilter/nf_conntrack_tcp_timeout_close_wait' to 3600
5月 19 12:15:56 localhost.localdomain kube-proxy[28949]: I0519 12:15:56.428522 28949 config.go:202] Starting service config controller
5月 19 12:15:56 localhost.localdomain kube-proxy[28949]: I0519 12:15:56.428558 28949 controller_utils.go:1019] Waiting for caches to sync for service config controller
5月 19 12:15:56 localhost.localdomain kube-proxy[28949]: I0519 12:15:56.428613 28949 config.go:102] Starting endpoints config controller
5月 19 12:15:56 localhost.localdomain kube-proxy[28949]: I0519 12:15:56.428622 28949 controller_utils.go:1019] Waiting for caches to sync for endpoints config controller
5月 19 12:15:56 localhost.localdomain kube-proxy[28949]: I0519 12:15:56.529703 28949 controller_utils.go:1026] Caches are synced for endpoints config controller
5月 19 12:15:56 localhost.localdomain kube-proxy[28949]: I0519 12:15:56.529834 28949 controller_utils.go:1026] Caches are synced for service config controller
5月 19 12:15:56 localhost.localdomain kube-proxy[28949]: I0519 12:15:56.530025 28949 proxier.go:329] Adding new service port "default/kubernetes:https" at 10.68.0.1:443/TCP
三、对搭建完成的k8s集群进行一些验证测试
在Master上查看k8s各项服务状态
[root@k8s-master bin]# kubectl get nodes
NAME STATUS ROLES AGE VERSION
10.0.2.4 Ready <none> 28m v1.9.8
10.0.2.5 Ready <none> 1m v1.9.8
[root@k8s-master bin]# kubectl get svc
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
kubernetes ClusterIP 10.10.10.1 <none> 443/TCP 3h
[root@k8s-master bin]# kubectl get cs
NAME STATUS MESSAGE ERROR
scheduler Healthy ok
controller-manager Healthy ok
etcd-0 Healthy {"health":"true"}
`
参考地址:https://blog.csdn.net/watermelonbig/article/details/80441654
https://www.cnblogs.com/liangyuntao-ts/p/10885352.html
`