一、准备环境

1.1 准备机器

准备三台CentOS7.2机器，一台作为master节点，其他作为node节点

修改主机名称

hostnamectl set-hostname k8s-mst

Role          IP                   Hostname

Master        192.168.0.87         k8s-cns1-mst

Node          192.168.0.88         k8s-cns1-nod1

Node          192.168.0.89         k8s-cns1-nod2

修改master节点/etc/hosts，添加以下内容（不设置的话，在master上kubectl相关命令无法操作对应主机上对象）

192.168.0.87     k8s-cns1-mst

192.168.0.88     k8s-cns1-nod1

192.168.0.89     k8s-cns1-nod2

为了避免和Docker的iptables产生冲突，关闭Node节点上的防火墙

systemctl stop firewalld

systemctl disable firewalld

了让各个节点的时间保持一致，所有节点安装NTP

yum -y install ntp

systemctl start ntpd

systemctl enable ntpd

1.2 安装docker（此处安装的是docker-ce版本）

[root@k8s-cns1-nod2 home]# cat installdocker.sh

#!/bin/bash

yum install -y yum-utils device-mapper-persistent-data lvm2

yum-config-manager --add-repo https://download.docker.com/linux/centos/docker-ce.repo

yum-config-manager --enable docker-ce-edge

yum-config-manager --enable docker-ce-testing

yum makecache fast

yum install -y docker-ce

运行Docker Daemon

systemctl start docker

二、源码编译

2.1 准备golang环境

参考https://golang.org/doc/install，下载对应版本并解压到/usr/local，例如

tar -C /usr/local -xzf go1.8.3.linux-amd64.tar.gz

安装git，下载kubenates源码并切换到需要分支

yum install git

go get -d k8s.io/kubernetes

cd /root/go/src/k8s.io/kubernetes

git checkout release-1.6.3      //使用release-1.6.3版本

make

编译成功后，可执行文件在

/root/go/src/k8s.io/kubernetes/_output/bin

三、Master配置

3.1 安装ectd（可选，如已有etcd集群略过）

3.1.1 软件安装

yum -y install etcd

3.1.2 修改etcd配置/etc/etcd/etcd.conf

ETCD_NAME=default

ETCD_DATA_DIR="/var/lib/etcd/default.etcd"

ETCD_LISTEN_CLIENT_URLS="http://0.0.0.0:2379"

3.1.3 运行etcd

systemctl enable etcd

systemctl start etcd

3.1.4 配置etcd子网

etcdctl mk /atomic.io/network/config '{"Network":"172.17.0.0/16"}'

3.2 kubernates配置

3.2.1 复制二进制

将位于/root/go/src/k8s.io/kubernetes/_output/bin/目录下的kube-apiserver、kube-controller-manager、kube-scheduler、kubectl复制到Master节点的/usr/bin/目录下

3.2.2 创建service配置脚本(shell)

[root@k8s-cns1-mst home]# cat configmaster.sh 
#!/bin/bash
# Copyright 2016 The Kubernetes Authors.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

MASTER_ADDRESS=${1:-"192.168.0.87"}
ETCD_SERVERS=${2:-"http://192.168.0.87:2379"}
SERVICE_CLUSTER_IP_RANGE=${3:-"10.254.0.0/16"}
ADMISSION_CONTROL=${4:-"NamespaceLifecycle,NamespaceExists,LimitRanger,SecurityContextDeny,ResourceQuota"}

cat <<EOF >/etc/kubernetes/config
# --logtostderr=true: log to standard error instead of files
KUBE_LOGTOSTDERR="--logtostderr=false"

# --v=0: log level for V logs
KUBE_LOG_LEVEL="--v=0"

# --allow-privileged=false: If true, allow privileged containers.
KUBE_ALLOW_PRIV="--allow-privileged=true"

# How the controller-manager, scheduler, and proxy find the apiserver
KUBE_MASTER="--master=${MASTER_ADDRESS}:8080"
EOF

cat <<EOF >/etc/kubernetes/apiserver
# --insecure-bind-address=127.0.0.1: The IP address on which to serve the --insecure-port.
KUBE_API_ADDRESS="--insecure-bind-address=0.0.0.0"

# --insecure-port=8080: The port on which to serve unsecured, unauthenticated access.
KUBE_API_PORT="--insecure-port=8080"

# --kubelet-port=10250: Kubelet port
NODE_PORT="--kubelet-port=10250"

# --etcd-servers=[]: List of etcd servers to watch (http://ip:port), 
# comma separated. Mutually exclusive with -etcd-config
KUBE_ETCD_SERVERS="--etcd-servers=${ETCD_SERVERS}"

# --advertise-address=<nil>: The IP address on which to advertise 
# the apiserver to members of the cluster.
KUBE_ADVERTISE_ADDR="--advertise-address=${MASTER_ADDRESS}"

# --service-cluster-ip-range=<nil>: A CIDR notation IP range from which to assign service cluster IPs. 
# This must not overlap with any IP ranges assigned to nodes for pods.
KUBE_SERVICE_ADDRESSES="--service-cluster-ip-range=${SERVICE_CLUSTER_IP_RANGE}"

# --admission-control="AlwaysAdmit": Ordered list of plug-ins 
# to do admission control of resources into cluster. 
# Comma-delimited list of: 
#   LimitRanger, AlwaysDeny, SecurityContextDeny, NamespaceExists, 
#   NamespaceLifecycle, NamespaceAutoProvision,
#   AlwaysAdmit, ServiceAccount, ResourceQuota, DefaultStorageClass
KUBE_ADMISSION_CONTROL="--admission-control=${ADMISSION_CONTROL}"

# Add your own!
KUBE_API_ARGS="--log-dir=/var/log/kubenates/"
EOF

KUBE_APISERVER_OPTS="   \${KUBE_LOGTOSTDERR}         \\
                        \${KUBE_LOG_LEVEL}           \\
                        \${KUBE_ETCD_SERVERS}        \\
                        \${KUBE_API_ADDRESS}         \\
                        \${KUBE_API_PORT}            \\
                        \${NODE_PORT}                \\
                        \${KUBE_ADVERTISE_ADDR}      \\
                        \${KUBE_ALLOW_PRIV}          \\
                        \${KUBE_SERVICE_ADDRESSES}   \\
                        \${KUBE_ADMISSION_CONTROL}   \\
                        \${KUBE_API_ARGS}"


cat <<EOF >/usr/lib/systemd/system/kube-apiserver.service
[Unit]
Description=Kubernetes API Server
Documentation=https://github.com/kubernetes/kubernetes
After=network.target
After=etcd.service

[Service]
EnvironmentFile=-/etc/kubernetes/config
EnvironmentFile=-/etc/kubernetes/apiserver
ExecStart=/usr/bin/kube-apiserver ${KUBE_APISERVER_OPTS}
Restart=on-failure
Type=notify
LimitNOFILE=65536

[Install]
WantedBy=multi-user.target
EOF

cat <<EOF >/etc/kubernetes/controller-manager
###
# The following values are used to configure the kubernetes controller-manager

# defaults from config and apiserver should be adequate

# Add your own!
KUBE_CONTROLLER_MANAGER_ARGS=""
EOF

KUBE_CONTROLLER_MANAGER_OPTS="  \${KUBE_LOGTOSTDERR} \\
                                \${KUBE_LOG_LEVEL}   \\
                                \${KUBE_MASTER}      \\
                                \${KUBE_CONTROLLER_MANAGER_ARGS}"

cat <<EOF >/usr/lib/systemd/system/kube-controller-manager.service
[Unit]
Description=Kubernetes Controller Manager
Documentation=https://github.com/kubernetes/kubernetes

[Service]
EnvironmentFile=-/etc/kubernetes/config
EnvironmentFile=-/etc/kubernetes/controller-manager
ExecStart=/usr/bin/kube-controller-manager ${KUBE_CONTROLLER_MANAGER_OPTS}
Restart=on-failure
LimitNOFILE=65536

[Install]
WantedBy=multi-user.target
EOF

cat <<EOF >/etc/kubernetes/scheduler
###
# kubernetes scheduler config

# Add your own!
KUBE_SCHEDULER_ARGS=""
EOF

KUBE_SCHEDULER_OPTS="   \${KUBE_LOGTOSTDERR}     \\
                        \${KUBE_LOG_LEVEL}       \\
                        \${KUBE_MASTER}          \\
                        \${KUBE_SCHEDULER_ARGS}"

cat <<EOF >/usr/lib/systemd/system/kube-scheduler.service
[Unit]
Description=Kubernetes Scheduler
Documentation=https://github.com/kubernetes/kubernetes

[Service]
EnvironmentFile=-/etc/kubernetes/config
EnvironmentFile=-/etc/kubernetes/scheduler
ExecStart=/usr/bin/kube-scheduler ${KUBE_SCHEDULER_OPTS}
Restart=on-failure
LimitNOFILE=65536

[Install]
WantedBy=multi-user.target
EOF

systemctl daemon-reload

3.2.3 创建service启动脚本

for svc in kube-apiserver kube-controller-manager kube-scheduler; do 
    systemctl restart $svc
    systemctl enable $svc
    systemctl status $svc 
done

四、Node配置

4.1 安装flanneld

4.1.1 安装软件

yum -y install flannel

4.1.2 修改flannel的配置文件/etc/sysconfig/flanneld

FLANNEL_ETCD="http://192.168.0.87:2379"
FLANNEL_ETCD_KEY="/atomic.io/network"

4.1.3 运行flannel

systemctl restart flanneld
systemctl enable flanneld
systemctl status flanneld

4.1.4 上传网络配置
创建一个config.json文件，内容如下：

{
"Network": "172.17.0.0/16",
"SubnetLen": 24,
"Backend": {
     "Type": "vxlan",
     "VNI": 7890
     }
 }

将配置上传到etcd服务器上

curl -L http://192.168.0.87:2379/v2/keys/atomic.io/network/config -XPUT --data-urlencode value@config.json

查看etcd分配的子网信息

[root@k8s-sz-0002 ~]# cat /run/flannel/subnet.env 
FLANNEL_NETWORK=172.17.0.0/16
FLANNEL_SUBNET=172.17.79.1/24
FLANNEL_MTU=1450
FLANNEL_IPMASQ=false

4.2 kubernates配置

4.2.1 复制二进制

将位于/root/go/src/k8s.io/kubernetes/_output/bin/目录下的kube-proxy、kubelet 复制到Node节点的/usr/bin/目录下

4.2.2 创建service配置脚本

[root@k8s-cns1-nod1 home]# cat configslave.sh 
#!/bin/bash
# Copyright 2016 The Kubernetes Authors.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

MASTER_ADDRESS=${1:-"192.168.0.87"}
NODE_HOSTNAME=${2:-"k8s-cns1-nod1"}

cat <<EOF >/etc/kubernetes/config
# --logtostderr=true: log to standard error instead of files
KUBE_LOGTOSTDERR="--logtostderr=true"

# --v=0: log level for V logs
KUBE_LOG_LEVEL="--v=0"

# --allow-privileged=false: If true, allow privileged containers.
KUBE_ALLOW_PRIV="--allow-privileged=false"

# How the controller-manager, scheduler, and proxy find the apiserver
KUBE_MASTER="--master=${MASTER_ADDRESS}:8080"
EOF

cat <<EOF >/etc/kubernetes/proxy
###
# kubernetes proxy config

# default config should be adequate

# Add your own!
KUBE_PROXY_ARGS="--log-dir=/var/log/kubenates/"
EOF

KUBE_PROXY_OPTS="   \${KUBE_LOGTOSTDERR} \\
                    \${KUBE_LOG_LEVEL}   \\
                    \${KUBE_MASTER}    \\
                    \${KUBE_PROXY_ARGS}"

cat <<EOF >/usr/lib/systemd/system/kube-proxy.service
[Unit]
Description=Kubernetes Proxy
After=network.target

[Service]
EnvironmentFile=-/etc/kubernetes/config
EnvironmentFile=-/etc/kubernetes/kube-proxy
ExecStart=/usr/bin/kube-proxy ${KUBE_PROXY_OPTS}
Restart=on-failure
LimitNOFILE=65536

[Install]
WantedBy=multi-user.target
EOF

cat <<EOF >/etc/kubernetes/kubelet
# --address=0.0.0.0: The IP address for the Kubelet to serve on (set to 0.0.0.0 for all interfaces)
KUBELET__ADDRESS="--address=0.0.0.0"

# --port=10250: The port for the Kubelet to serve on. Note that "kubectl logs" will not work if you set this flag.
KUBELET_PORT="--port=10250"

# --hostname-override="": If non-empty, will use this string as identification instead of the actual hostname.
KUBELET_HOSTNAME="--hostname-override=${NODE_HOSTNAME}"

# --api-servers=[]: List of Kubernetes API servers for publishing events, 
# and reading pods and services. (ip:port), comma separated.
KUBELET_API_SERVER="--api-servers=http://${MASTER_ADDRESS}: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=""
EOF

KUBE_PROXY_OPTS="   \${KUBE_LOGTOSTDERR}     \\
                    \${KUBE_LOG_LEVEL}       \\
                    \${KUBELET__ADDRESS}         \\
                    \${KUBELET_PORT}            \\
                    \${KUBELET_HOSTNAME}        \\
                    \${KUBELET_API_SERVER}   \\
                    \${KUBE_ALLOW_PRIV}      \\
                    \${KUBELET_POD_INFRA_CONTAINER}\\
                    \${KUBELET_ARGS}"

cat <<EOF >/usr/lib/systemd/system/kubelet.service
[Unit]
Description=Kubernetes Kubelet
After=docker.service
Requires=docker.service

[Service]
WorkingDirectory=/var/lib/kubelet
EnvironmentFile=-/etc/kubernetes/config
EnvironmentFile=-/etc/kubernetes/kubelet
ExecStart=/usr/bin/kubelet ${KUBE_PROXY_OPTS}
Restart=on-failure

[Install]
WantedBy=multi-user.target
EOF

systemctl daemon-reload

4.2.3 创建service启动脚本

# cat /home/startslave.sh 
mkdir -p /etc/kubernetes/
mkdir -p /var/logs/kubernetes/
mkdir -p /var/lib/kubelet

source /run/flannel/subnet.env

sed -i "s|--bip=.*|--bip=${FLANNEL_SUBNET} --ip-masq=true --mtu=1472|" /usr/lib/systemd/system/docker.service

systemctl daemon-reload

for svc in docker kubelet kube-proxy; do 
    systemctl restart $svc
    systemctl enable $svc
    systemctl status $svc 
done

必须注意，/usr/lib/systemd/system/docker.service是docker 的配置文件，因为node和etcd之间是有租约时间的，如果node长时间不在线的话，etcd会认为租约超期，清除子网信息，当node起来后，flannel重新获取新的子网，所以docker容器网段(--bip=)建议每次都从新的flannel获取，保证一致

编辑/etc/rc.local，将脚本执行加入开机启动项，保证node重启后自动执行脚本

验证环境配置

在Master节点运行命令kubectl get nodes，输出信息如下：

[root@k8s-cns1-mst home]# kubectl get node
NAME            STATUS    AGE       VERSION
k8s-cns1-nod1   Ready     1d        v1.6.4
k8s-cns1-nod2   Ready     1d        v1.6.4

常见问题

1、使用etcdctl时连接本地报错

/usr/lib # etcdctl ls /
Error:  client: etcd cluster is unavailable or misconfigured; error #0: dial tcp 127.0.0.1:2379: getsockopt: connection refused
; error #1: dial tcp 127.0.0.1:4001: getsockopt: connection refused

error #0: dial tcp 127.0.0.1:2379: getsockopt: connection refused
error #1: dial tcp 127.0.0.1:4001: getsockopt: connection refused

这是因为ETCD_LISTEN_CLIENT_URLS参数没有配置http://127.0.0.1:2379而导致，使用etcdctl时加上endpoints选项

/usr/lib # etcdctl --endpoints=192.168.0.87:2379 ls /