Kubernetes集群部署篇( 一)
K8S集群部署有几种方式:kubeadm、minikube和二进制包。前两者属于自动部署,简化部署操作,我们这里强烈推荐初学者使用二进制包部署,因为自动部署屏蔽了很多细节,使得对各个模块感知很少,非常不利用学习。所以,这篇文章也是使用二进制包部署Kubernetes集群。
一、架构拓扑图
二、环境规划
角色 | IP | 主机名 | 组件 |
---|---|---|---|
Master1 | 192.168.161.161 | master1 | etcd1,master1 |
master2 | 192.168.161.162 | master2 | etcd2,master2 |
node1 | 192.168.161.163 | node1 | kubelet,kube-proxy,docker,flannel |
node2 | 192.168.161.164 | node2 | kubelet,kube-proxy,docker,flannel |
- kube-apiserver:位于master节点,接受用户请求。
- kube-scheduler:位于master节点,负责资源调度,即pod建在哪个node节点。
- kube-controller-manager:位于master节点,包含ReplicationManager,Endpointscontroller,Namespacecontroller,Nodecontroller等。
- etcd:分布式键值存储系统,共享整个集群的资源对象信息。
- kubelet:位于node节点,负责维护在特定主机上运行的pod。
- kube-proxy:位于node节点,它起的作用是一个服务代理的角色。
再来普及一下:
① kubectl 发送部署请求到 API Server。
② API Server 通知 Controller Manager 创建一个 deployment 资源。
③ Scheduler 执行调度任务,将两个副本 Pod 分发到 k8s-node1 和 k8s-node2。
④ k8s-node1 和 k8s-node2 上的 kubectl 在各自的节点上创建并运行 Pod。
三、集群部署
- 系统采用 Centos 7.3
- 关闭防火墙
systemctl disable firewalld
systemctl stop firewalld
- 关闭 selinux
- 安装NTP并启动
# yum -y install ntp
# systemctl start ntpd
# systemctl enable ntpd
4台机器均设置好 hosts
vim /etc/hosts
192.168.161.161 master1
192.168.161.162 master2
192.168.161.163 node1
192.168.161.164 node2
192.168.161.161 etcd
192.168.161.162 etcd
3.1 部署master
安装etcd
[root@master1 ~]# yum -y install etcd
配置etcd
yum安装的etcd默认配置文件在/etc/etcd/etcd.conf,以下将2个节点上的配置贴出来,请注意不同点。
2379是默认的使用端口,为了防止端口占用问题的出现,增加4001端口备用。
master1:
[root@master1 ~]# vim /etc/etcd/etcd.conf
# [member]
ETCD_NAME=etcd1
ETCD_DATA_DIR="/var/lib/etcd/test.etcd"
#ETCD_WAL_DIR=""
#ETCD_SNAPSHOT_COUNT="10000"
#ETCD_HEARTBEAT_INTERVAL="100"
#ETCD_ELECTION_TIMEOUT="1000"
ETCD_LISTEN_PEER_URLS="http://0.0.0.0:2380"
ETCD_LISTEN_CLIENT_URLS="http://0.0.0.0:2379,http://0.0.0.0:4001"
#ETCD_MAX_SNAPSHOTS="5"
#ETCD_MAX_WALS="5"
#ETCD_CORS=""
#
#[cluster]
ETCD_INITIAL_ADVERTISE_PEER_URLS="http://master1:2380"
# if you use different ETCD_NAME (e.g. test), set ETCD_INITIAL_CLUSTER value for this name, i.e. "test=http://..."
ETCD_INITIAL_CLUSTER="etcd1=http://master1:2380,etcd2=http://master2:2380"
ETCD_INITIAL_CLUSTER_STATE="new"
ETCD_INITIAL_CLUSTER_TOKEN="etcd-cluster-baby"
ETCD_ADVERTISE_CLIENT_URLS="http://master1:2379,http://master1:4001"
master2:
[root@master2 ~]# vim /etc/etcd/etcd.conf
# [member]
ETCD_NAME=etcd2
ETCD_DATA_DIR="/var/lib/etcd/test.etcd"
#ETCD_WAL_DIR=""
#ETCD_SNAPSHOT_COUNT="10000"
#ETCD_HEARTBEAT_INTERVAL="100"
#ETCD_ELECTION_TIMEOUT="1000"
ETCD_LISTEN_PEER_URLS="http://0.0.0.0:2380"
ETCD_LISTEN_CLIENT_URLS="http://0.0.0.0:2379,http://0.0.0.0:4001"
#ETCD_MAX_SNAPSHOTS="5"
#ETCD_MAX_WALS="5"
#ETCD_CORS=""
#
#[cluster]
ETCD_INITIAL_ADVERTISE_PEER_URLS="http://master2:2380"
# if you use different ETCD_NAME (e.g. test), set ETCD_INITIAL_CLUSTER value for this name, i.e. "test=http://..."
ETCD_INITIAL_CLUSTER="etcd1=http://master1:2380,etcd2=http://master2:2380"
ETCD_INITIAL_CLUSTER_STATE="new"
ETCD_INITIAL_CLUSTER_TOKEN="etcd-cluster-baby"
ETCD_ADVERTISE_CLIENT_URLS="http://master2:2379,http://master2:4001"
参数说明:
name 节点名称
data-dir 指定节点的数据存储目录
listen-peer-urls 监听URL,用于与其他节点通讯
listen-client-urls 对外提供服务的地址:比如 http://ip:2379,http://127.0.0.1:2379 ,客户端会连接到这里和 etcd 交互
initial-advertise-peer-urls 该节点同伴监听地址,这个值会告诉集群中其他节点
initial-cluster 集群中所有节点的信息,格式为 node1=http://ip1:2380,node2=http://ip2:2380,… 。注意:这里的 node1 是节点的 --name 指定的名字;后面的 ip1:2380 是 --initial-advertise-peer-urls 指定的值
initial-cluster-state 新建集群的时候,这个值为 new ;假如已经存在的集群,这个值为 existing
initial-cluster-token 创建集群的 token,这个值每个集群保持唯一。这样的话,如果你要重新创建集群,即使配置和之前一样,也会再次生成新的集群和节点 uuid;否则会导致多个集群之间的冲突,造成未知的错误
advertise-client-urls 对外公告的该节点客户端监听地址,这个值会告诉集群中其他节点
修改好以上配置后,在各个节点上启动etcd服务,并验证集群状态:
Master1
[root@master1 etcd]# systemctl start etcd
[root@master1 etcd]# etcdctl -C http://etcd:2379 cluster-health
member 22a9f7f65563bff5 is healthy: got healthy result from http://master2:2379
member d03b92adc5af7320 is healthy: got healthy result from http://master1:2379
cluster is healthy
[root@master1 etcd]# etcdctl -C http://etcd:4001 cluster-health
member 22a9f7f65563bff5 is healthy: got healthy result from http://master2:2379
member d03b92adc5af7320 is healthy: got healthy result from http://master1:2379
cluster is healthy
[root@master1 etcd]# systemctl enable etcd
Created symlink from /etc/systemd/system/multi-user.target.wants/etcd.service to /usr/lib/systemd/system/etcd.service.
Master2
[root@master2 etcd]# systemctl start etcd
[root@master2 etcd]# etcdctl -C http://etcd:2379 cluster-health
member 22a9f7f65563bff5 is healthy: got healthy result from http://master2:2379
member d03b92adc5af7320 is healthy: got healthy result from http://master1:2379
cluster is healthy
[root@master2 etcd]# etcdctl -C http://etcd:4001 cluster-health
member 22a9f7f65563bff5 is healthy: got healthy result from http://master2:2379
member d03b92adc5af7320 is healthy: got healthy result from http://master1:2379
cluster is healthy
[root@master2 etcd]# systemctl enable etcd
Created symlink from /etc/systemd/system/multi-user.target.wants/etcd.service to /usr/lib/systemd/system/etcd.service.
部署 master
安装 docker ,设置开机自启动并开启服务
分别在 master1和master2上面安装docker服务
[root@master1 etcd]# yum install docker -y
[root@master1 etcd]# chkconfig docker on
[root@master1 etcd]# systemctl start docker.service
安装kubernets
yum install kubernetes -y
在 master 的虚机上,需要运行三个组件:Kubernets API Server、Kubernets Controller Manager、Kubernets Scheduler。
首先修改 /etc/kubernetes/apiserver 文件:
[root@master1 kubernetes]# vim apiserver
###
# kubernetes system config
#
# The following values are used to configure the kube-apiserver
#
# The address on the local server to listen to.
KUBE_API_ADDRESS="--insecure-bind-address=0.0.0.0"
# The port on the local server to listen on.
KUBE_API_PORT="--port=8080"
# Port minions listen on
# KUBELET_PORT="--kubelet-port=10250"
# Comma separated list of nodes in the etcd cluster
KUBE_ETCD_SERVERS="--etcd-servers=http://etcd:2379"
# Address range to use for services
KUBE_SERVICE_ADDRESSES="--service-cluster-ip-range=10.254.0.0/16"
# default admission control policies
# KUBE_ADMISSION_CONTROL="--admission-control=NamespaceLifecycle,NamespaceExists,LimitRanger,SecurityContextDeny,ServiceAccount,ResourceQuota"
KUBE_ADMISSION_CONTROL="--admission-control=NamespaceLifecycle,NamespaceExists,LimitRanger,SecurityContextDeny,ResourceQuota"
# Add your own!
KUBE_API_ARGS=""
接着修改 /etc/kubernetes/config 文件:(最后一句 masterX:8080 ,对应master1/2机器就好)
[root@master1 ~]# vim /etc/kubernetes/config
###
# kubernetes system config
#
# The following values are used to configure various aspects of all
# kubernetes services, including
#
# kube-apiserver.service
# kube-controller-manager.service
# kube-scheduler.service
# kubelet.service
# kube-proxy.service
# logging to stderr means we get it in the systemd journal
KUBE_LOGTOSTDERR="--logtostderr=true"
# journal message level, 0 is debug
KUBE_LOG_LEVEL="--v=0"
# Should this cluster be allowed to run privileged docker containers
KUBE_ALLOW_PRIV="--allow-privileged=false"
# How the controller-manager, scheduler, and proxy find the apiserver
KUBE_MASTER="--master=http://master1:8080"
修改完成后,启动服务并设置开机自启动即可:
systemctl enable kube-apiserver
systemctl start kube-apiserver
systemctl enable kube-controller-manager
systemctl start kube-controller-manager
systemctl enable kube-scheduler
systemctl start kube-scheduler
部署 node
安装 docker ,设置开机自启动并开启服务
yum install docker -y
chkconfig docker on
systemctl start docker.service
安装 kubernetes
yum install kubernetes -y
在 node 的虚机上,需要运行三个组件:Kubelet、Kubernets Proxy。
首先修改 /etc/kubernetes/config 文件:(注意:这里配置的是etcd的地址,也就是master1/2的地址其中之一)
[root@node1 ~]# vim /etc/kubernetes/config
###
# kubernetes system config
#
# The following values are used to configure various aspects of all
# kubernetes services, including
#
# kube-apiserver.service
# kube-controller-manager.service
# kube-scheduler.service
# kubelet.service
# kube-proxy.service
# logging to stderr means we get it in the systemd journal
KUBE_LOGTOSTDERR="--logtostderr=true"
# journal message level, 0 is debug
KUBE_LOG_LEVEL="--v=0"
# Should this cluster be allowed to run privileged docker containers
KUBE_ALLOW_PRIV="--allow-privileged=false"
# How the controller-manager, scheduler, and proxy find the apiserver
KUBE_MASTER="--master=http://etcd:8080"
接着修改 /etc/kubernetes/kubelet 文件:(注:–hostname-override= 对应的node机器)
[root@node1 ~]# vim /etc/kubernetes/kubelet
###
# kubernetes kubelet (minion) config
# The address for the info server to serve on (set to 0.0.0.0 or "" for all interfaces)
KUBELET_ADDRESS="--address=0.0.0.0"
# The port for the info server to serve on
# KUBELET_PORT="--port=10250"
# You may leave this blank to use the actual hostname
KUBELET_HOSTNAME="--hostname-override=node1"
# location of the api-server
KUBELET_API_SERVER="--api-servers=http://etcd:8080"
# pod infrastructure container
KUBELET_POD_INFRA_CONTAINER="--pod-infra-container-image=registry.access.redhat.com/rhel7/pod-infrastructure:latest"
# Add your own!
KUBELET_ARGS=""
修改完成后,启动服务并设置开机自启动即可:
systemctl enable kubelet
systemctl start kubelet
systemctl enable kube-proxy
systemctl start kube-proxy
查看集群状态
在任意一台master上查看集群中节点及节点状态:
[root@master1 kubernetes]# kubectl get node
NAME STATUS AGE
node1 Ready 1m
node2 Ready 1m
至此,已经搭建了一个kubernetes集群了,但目前该集群还不能很好的工作,因为需要对集群中pod的网络进行统一管理。
创建覆盖网络 flannel
在master、node上均执行如下命令,安装 flannel
yum install flannel -y
在master、node上均编辑 /etc/sysconfig/flanneld 文件
[root@master1 kubernetes]# vim /etc/sysconfig/flanneld
# Flanneld configuration options
# etcd url location. Point this to the server where etcd runs
FLANNEL_ETCD_ENDPOINTS="http://etcd:2379"
# etcd config key. This is the configuration key that flannel queries
# For address range assignment
FLANNEL_ETCD_PREFIX="/atomic.io/network"
# Any additional options that you want to pass
#FLANNEL_OPTIONS=""
flannel使用etcd进行配置,来保证多个flannel实例之间的配置一致性,所以需要在etcd上进行如下配置:
etcdctl mk /atomic.io/network/config '{ "Network": "10.0.0.0/16" }'
(‘/atomic.io/network/config’这个key与上文/etc/sysconfig/flannel中的配置项FLANNEL_ETCD_PREFIX是相对应的,错误的话启动就会出错)
启动修改后的 flannel ,并依次重启docker、kubernete
在 master 虚机上执行:
systemctl enable flanneld
systemctl start flanneld
service docker restart
systemctl restart kube-apiserver
systemctl restart kube-controller-manager
systemctl restart kube-scheduler
在 node 虚机上执行:
systemctl enable flanneld
systemctl start flanneld
service docker restart
systemctl restart kubelet
systemctl restart kube-proxy
这样etcd集群 + flannel + kubernetes集群 在centOS7上就搭建起来了。
注:
flannel架构介绍
flannel默认使用8285端口作为UDP封装报文的端口,VxLan使用8472端口。
那么一条网络报文是怎么从一个容器发送到另外一个容器的呢?
1. 容器直接使用目标容器的ip访问,默认通过容器内部的eth0发送出去。
2. 报文通过veth pair被发送到vethXXX。
3. vethXXX是直接连接到虚拟交换机docker0的,报文通过虚拟bridge docker0发送出去。
4. 查找路由表,外部容器ip的报文都会转发到flannel0虚拟网卡,这是一个P2P的虚拟网卡,然后报文就被转发到监听在另一端的flanneld。
5. flanneld通过etcd维护了各个节点之间的路由表,把原来的报文UDP封装一层,通过配置的iface发送出去。
6. 报文通过主机之间的网络找到目标主机。
7. 报文继续往上,到传输层,交给监听在8285端口的flanneld程序处理。
8. 数据被解包,然后发送给flannel0虚拟网卡。
9. 查找路由表,发现对应容器的报文要交给docker0。
10. docker0找到连到自己的容器,把报文发送过去。