k8s二进制部署

k8s二进制部署

 

1.环境准备

主机名	ip地址	角色
k8s-master01	10.0.0.10	master
k8s-master02	10.0.0.11	master
k8s-node01	10.0.0.12	node
k8s-node02	10.0.0.13	node

初始化操作

• 关闭防火墙

• 关闭selinux

• 关闭swap

• 安装ntp使时间同步

• 配置域名解析

• 配置免密 k8s-master01 到其他机器。

• 安装docker

2.生成配置CFSSL

CFFSL能够构建本地CA,生成后面需要使用的证书

再k8s-master01上操作

wget https://pkg.cfssl.org/R1.2/cfssl_linux-amd64
chmod +x cfssl_linux-amd64
sudo mv cfssl_linux-amd64 /root/local/bin/cfssl
​
wget https://pkg.cfssl.org/R1.2/cfssljson_linux-amd64
chmod +x cfssljson_linux-amd64
sudo mv cfssljson_linux-amd64 /root/local/bin/cfssljson
​
wget https://pkg.cfssl.org/R1.2/cfssl-certinfo_linux-amd64
chmod +x cfssl-certinfo_linux-amd64
sudo mv cfssl-certinfo_linux-amd64 /root/local/bin/cfssl-certinfo
​
export PATH=/root/local/bin:$PATH

3.配置ansible

编辑k8s-master01 机器上的/etc/ansible/hosts 加入以下内容

k8s-master01
k8s-master02
k8s-node01
k8s-node02
[master]
k8s-master01
k8s-master02
[node]
k8s-node01
k8s-node02
[other]
k8s-master02
k8s-node01
k8s-node02

4.升级内核

因为3.10版本内核且缺少 ip_vs_fo.ko 模块，将导致 kube-proxy 无法开启ipvs模式。ip_vs_fo.ko 模块的最早版本为3.19版本，这个内核版本在 RedHat 系列发行版的常见RPM源中是不存在的。

[root@k8s-master01 ~]# ansible all -a 'rpm -Uvh http://www.elrepo.org/elrepo-release-7.0-2.el7.elrepo.noarch.rpm'
​
[root@k8s-master01 ~]# ansible all -a 'yum --enablerepo=elrepo-kernel install kernel-ml-devel kernel-ml -y'
​
[root@k8s-master01 ~]# ansible all -a 'uname -r'
​

安装好之后重启，选择升级的后的内核加载

[root@k8s-master01 ~]# ansible all -a 'modprobe ip_vs_fo'
[root@k8s-master01 ~]# ansible all -m raw -a "lsmod|grep vs"

 

5.创建安装目录

[root@k8s-master01 ~]# ansible all -m raw -a 'mkdir /home/work/_app/k8s/etcd/{bin,cfg,ssl} -p'
​
[root@k8s-master01 ~]# ansible all -a 'ls /home/work/_app/k8s/etcd/'

6.创建etcd证书

[root@k8s-master01 ~]#  mkdir -p /home/work/_src/ssl_etcd
[root@k8s-master01 ~]#  cd /home/work/_src/ssl_etcd

6.1生成 ETCD SERVER 证书用到的JSON请求文件

[root@k8s-master01 ssl_etcd]# cat << EOF | tee ca-config.json
{
  "signing": {
    "default": {
      "expiry": "87600h"
   },
    "profiles": {
      "etcd": {
         "expiry": "87600h",
         "usages": [
            "signing",
            "key encipherment",
            "server auth",
            "client auth"
       ]
     }
   }
 }
}
EOF
## 说明 生成ca-config.json
默认策略，指定了证书的有效期是10年(87600h)
etcd策略，指定了证书的用途
signing, 表示该证书可用于签名其它证书；生成的 ca.pem 证书中 CA=TRUE
server auth：表示 client 可以用该 CA 对 server 提供的证书进行验证
client auth：表示 server 可以用该 CA 对 client 提供的证书进行验证

6.2创建 ETCD CA 证书配置文件

[root@k8s-master01 ssl_etcd]# cat << EOF | tee ca-csr.json
{
    "CN": "etcd CA",
    "key": {
        "algo": "rsa",
        "size": 2048
   },
    "names": [
       {
            "C": "CN",
            "L": "Beijing",
            "ST": "Beijing"
       }
   ]
}
EOF
​
#生成ca-csr.json

6.3 创建 ETCD SERVER 证书配置文件

[root@k8s-master01 ssl_etcd]# cat << EOF | tee server-csr.json
{
    "CN": "etcd",
    "hosts": [
    "10.0.0.10",
    "10.0.0.11",
    "10.0.0.12",
    "10.0.0.13"
    ],
    "key": {
        "algo": "rsa",
        "size": 2048
    },
    "names": [
        {
            "C": "CN",
            "L": "Beijing",
            "ST": "Beijing"
        }
    ]
}
EOF

#生成server-csr.json

6.4 生成ETCD CA证书和私钥

[root@k8s-master01 ssl_etcd]# cfssl gencert -initca ca-csr.json | cfssljson -bare ca
2019/10/29 15:19:05 [INFO] generating a new CA key and certificate from CSR
2019/10/29 15:19:05 [INFO] generate received request
2019/10/29 15:19:05 [INFO] received CSR
2019/10/29 15:19:05 [INFO] generating key: rsa-2048
2019/10/29 15:19:06 [INFO] encoded CSR
2019/10/29 15:19:06 [INFO] signed certificate with serial number 118073519875290282867413793117201542018807809668

#生成 ca.csr ca-key.pem  ca.pem

6.5 生成 ETCD SERVER 证书和私钥

[root@k8s-master01 ssl_etcd]# cfssl gencert -ca=ca.pem -ca-key=ca-key.pem -config=ca-config.json -profile=etcd server-csr.json | cfssljson -bare server
2019/10/29 15:20:40 [INFO] generate received request
2019/10/29 15:20:40 [INFO] received CSR
2019/10/29 15:20:40 [INFO] generating key: rsa-2048
2019/10/29 15:20:40 [INFO] encoded CSR
2019/10/29 15:20:40 [INFO] signed certificate with serial number 567338985519564573573112117172193680391181260406
2019/10/29 15:20:40 [WARNING] This certificate lacks a "hosts" field. This makes it unsuitable for
websites. For more information see the Baseline Requirements for the Issuance and Management
of Publicly-Trusted Certificates, v.1.1.6, from the CA/Browser Forum (https://cabforum.org);
specifically, section 10.2.3 ("Information Requirements").
[root@k8s-master01 ssl_etcd]# ls
ca-config.json  ca.csr  ca-csr.json  ca-key.pem  ca.pem  server.csr  server-csr.json  server-key.pem  server.pem

# 生成 server.csr server-key.pem  server.pem

• 将生成的证书，复制到证书目录

[root@k8s-master01 ssl_etcd]# cp *.pem /home/work/_app/k8s/etcd/ssl/

7.安装ETCD

7.1下载ETCD

[root@k8s-master01 _src]# cd /home/work/_src
[root@k8s-master01 _src]# wget https://github.com/etcd-io/etcd/releases/download/v3.3.10/etcd-v3.3.10-linux-amd64.tar.gz
[root@k8s-master01 _src]# tar -xvf etcd-v3.3.10-linux-amd64.tar.gz 
[root@k8s-master01 _src]# cd etcd-v3.3.10-linux-amd64
[root@k8s-master01 etcd-v3.3.10-linux-amd64]# cp etcd etcdctl /home/work/_app/k8s/etcd/bin/

7.2创建ETCD系统启动文件

创建 /usr/lib/systemd/system/etcd.service 文件并保存，内容如下：

[root@k8s-master01 _src]# cat /usr/lib/systemd/system/etcd.service 
[Unit]
Description=Etcd Server
After=network.target
After=network-online.target
Wants=network-online.target

[Service]
Type=notify
EnvironmentFile=/home/work/_app/k8s/etcd/cfg/etcd.conf
ExecStart=/home/work/_app/k8s/etcd/bin/etcd \
--name=${ETCD_NAME} \
--data-dir=${ETCD_DATA_DIR} \
--listen-peer-urls=${ETCD_LISTEN_PEER_URLS} \
--listen-client-urls=${ETCD_LISTEN_CLIENT_URLS},http://127.0.0.1:2379 \
--advertise-client-urls=${ETCD_ADVERTISE_CLIENT_URLS} \
--initial-advertise-peer-urls=${ETCD_INITIAL_ADVERTISE_PEER_URLS} \
--initial-cluster=${ETCD_INITIAL_CLUSTER} \
--initial-cluster-token=${ETCD_INITIAL_CLUSTER_TOKEN} \
--initial-cluster-state=new \
--cert-file=/home/work/_app/k8s/etcd/ssl/server.pem \
--key-file=/home/work/_app/k8s/etcd/ssl/server-key.pem \
--peer-cert-file=/home/work/_app/k8s/etcd/ssl/server.pem \
--peer-key-file=/home/work/_app/k8s/etcd/ssl/server-key.pem \
--trusted-ca-file=/home/work/_app/k8s/etcd/ssl/ca.pem \
--peer-trusted-ca-file=/home/work/_app/k8s/etcd/ssl/ca.pem
Restart=on-failure
LimitNOFILE=65536

[Install]
WantedBy=multi-user.target

7.3 将 ETCD 启动文件、证书文件、系统启动文件分发到其他节点

[root@k8s-master01 _app]# ansible other -m copy -a "src=/home/work/_app/k8s/etcd dest=/home/work/_app/k8s/ mode=0744"
[root@k8s-master01 _app]# ansible other -a "ls /home/work/_app/k8s/etcd"

[root@k8s-master01 _app]# ansible other -m copy -a "src=/usr/lib/systemd/system/etcd.service dest=/usr/lib/systemd/system/"
[root@k8s-master01 _app]# ansible other -a "ls /usr/lib/systemd/system/etcd.service"

7.4 创建ETCD主配置文件

 在 k8s-master01 创建 /home/work/_app/k8s/etcd/cfg/etcd.conf 文件，内容如下：

[root@k8s-master01 _src]# cat << EOF | tee /home/work/_app/k8s/etcd/cfg/etcd.conf
#[Member]
# ETCD的节点名
ETCD_NAME="etcd00"
# ETCD的数据存储目录
ETCD_DATA_DIR="/home/work/_data/etcd"
# 该节点与其他节点通信时所监听的地址列表，多个地址使用逗号隔开，其格式可以划分为scheme://IP:PORT，这里的scheme可以是http、https
ETCD_LISTEN_PEER_URLS="https://10.0.0.10:2380"
# 该节点与客户端通信时监听的地址列表
ETCD_LISTEN_CLIENT_URLS="https://10.0.0.10:2379"
 
#[Clustering]
# 该成员节点在整个集群中的通信地址列表，这个地址用来传输集群数据的地址。因此这个地址必须是可以连接集群中所有的成员的。
ETCD_INITIAL_ADVERTISE_PEER_URLS="https://10.0.0.10:2380"
# 配置集群内部所有成员地址，其格式为：ETCD_NAME=ETCD_INITIAL_ADVERTISE_PEER_URLS，如果有多个使用逗号隔开
ETCD_ADVERTISE_CLIENT_URLS="https://10.0.0.10:2379"
ETCD_INITIAL_CLUSTER="etcd00=https://10.0.0.10:2380,etcd01=https://10.0.0.11:2380,etcd02=https://10.0.0.12:2380,etcd03=https://10.0.0.13:2380"
# 初始化集群token
ETCD_INITIAL_CLUSTER_TOKEN="etcd-cluster"
# 初始化集群状态，new表示新建
ETCD_INITIAL_CLUSTER_STATE="new"

#[Security]
ETCD_CERT_FILE="/home/work/_app/k8s/etcd/ssl/server.pem"
ETCD_KEY_FILE="/home/work/_app/k8s/etcd/ssl/server-key.pem"
ETCD_TRUSTED_CA_FILE="/home/work/_app/k8s/etcd/ssl/ca.pem"
ETCD_CLIENT_CERT_AUTH="true"
ETCD_PEER_CERT_FILE="/home/work/_app/k8s/etcd/ssl/server.pem"
ETCD_PEER_KEY_FILE="/home/work/_app/k8s/etcd/ssl/server-key.pem"
ETCD_PEER_TRUSTED_CA_FILE="/home/work/_app/k8s/etcd/ssl/ca.pem"
ETCD_PEER_CLIENT_CERT_AUTH="true"

EOF

在 k8s-master02 创建 /home/work/_app/k8s/etcd/cfg/etcd.conf 文件，内容如下：

[root@k8s-master02 _src]# cat << EOF | tee /home/work/_app/k8s/etcd/cfg/etcd.conf
#[Member]
ETCD_NAME="etcd01"
ETCD_DATA_DIR="/home/work/_data/etcd"
ETCD_LISTEN_PEER_URLS="https://10.0.0.11:2380"
ETCD_LISTEN_CLIENT_URLS="https://10.0.0.11:2379"
 
#[Clustering]
ETCD_INITIAL_ADVERTISE_PEER_URLS="https://10.0.0.11:2380"
ETCD_ADVERTISE_CLIENT_URLS="https://10.0.0.11:2379"
ETCD_INITIAL_CLUSTER="etcd00=https://10.0.0.10:2380,etcd01=https://10.0.0.11:2380,etcd02=https://10.0.0.12:2380,etcd03=https://10.0.0.13:2380"
ETCD_INITIAL_CLUSTER_TOKEN="etcd-cluster"
ETCD_INITIAL_CLUSTER_STATE="new"

#[Security]
ETCD_CERT_FILE="/home/work/_app/k8s/etcd/ssl/server.pem"
ETCD_KEY_FILE="/home/work/_app/k8s/etcd/ssl/server-key.pem"
ETCD_TRUSTED_CA_FILE="/home/work/_app/k8s/etcd/ssl/ca.pem"
ETCD_CLIENT_CERT_AUTH="true"
ETCD_PEER_CERT_FILE="/home/work/_app/k8s/etcd/ssl/server.pem"
ETCD_PEER_KEY_FILE="/home/work/_app/k8s/etcd/ssl/server-key.pem"
ETCD_PEER_TRUSTED_CA_FILE="/home/work/_app/k8s/etcd/ssl/ca.pem"
ETCD_PEER_CLIENT_CERT_AUTH="true"

EOF

 在 k8s-node01 创建 /home/work/_app/k8s/etcd/cfg/etcd.conf 文件，内容如下：

[root@k8s-node01 home]# cat << EOF | tee /home/work/_app/k8s/etcd/cfg/etcd.conf
#[Member]
ETCD_NAME="etcd02"
ETCD_DATA_DIR="/home/work/_data/etcd"
ETCD_LISTEN_PEER_URLS="https://10.0.0.12:2380"
ETCD_LISTEN_CLIENT_URLS="https://10.0.0.12:2379"
 
#[Clustering]
ETCD_INITIAL_ADVERTISE_PEER_URLS="https://10.0.0.12:2380"
ETCD_ADVERTISE_CLIENT_URLS="https://10.0.0.12:2379"
ETCD_INITIAL_CLUSTER="etcd00=https://10.0.0.10:2380,etcd01=https://10.0.0.11:2380,etcd02=https://10.0.0.12:2380,etcd03=https://10.0.0.13:2380"
ETCD_INITIAL_CLUSTER_TOKEN="etcd-cluster"
ETCD_INITIAL_CLUSTER_STATE="new"

#[Security]
ETCD_CERT_FILE="/home/work/_app/k8s/etcd/ssl/server.pem"
ETCD_KEY_FILE="/home/work/_app/k8s/etcd/ssl/server-key.pem"
ETCD_TRUSTED_CA_FILE="/home/work/_app/k8s/etcd/ssl/ca.pem"
ETCD_CLIENT_CERT_AUTH="true"
ETCD_PEER_CERT_FILE="/home/work/_app/k8s/etcd/ssl/server.pem"
ETCD_PEER_KEY_FILE="/home/work/_app/k8s/etcd/ssl/server-key.pem"
ETCD_PEER_TRUSTED_CA_FILE="/home/work/_app/k8s/etcd/ssl/ca.pem"
ETCD_PEER_CLIENT_CERT_AUTH="true"

EOF

在 k8s-node02 创建 /home/work/_app/k8s/etcd/cfg/etcd.conf 文件，内容如下：

[root@k8s-node02 home]#  cat << EOF | tee /home/work/_app/k8s/etcd/cfg/etcd.conf
#[Member]
ETCD_NAME="etcd03"
ETCD_DATA_DIR="/home/work/_data/etcd"
ETCD_LISTEN_PEER_URLS="https://10.0.0.13:2380"
ETCD_LISTEN_CLIENT_URLS="https://10.0.0.13:2379"
 
#[Clustering]
ETCD_INITIAL_ADVERTISE_PEER_URLS="https://10.0.0.13:2380"
ETCD_ADVERTISE_CLIENT_URLS="https://10.0.0.13:2379"
ETCD_INITIAL_CLUSTER="etcd00=https://10.0.0.10:2380,etcd01=https://10.0.0.11:2380,etcd02=https://10.0.0.12:2380,etcd03=https://10.0.0.13:2380"
ETCD_INITIAL_CLUSTER_TOKEN="etcd-cluster"
ETCD_INITIAL_CLUSTER_STATE="new"

#[Security]
ETCD_CERT_FILE="/home/work/_app/k8s/etcd/ssl/server.pem"
ETCD_KEY_FILE="/home/work/_app/k8s/etcd/ssl/server-key.pem"
ETCD_TRUSTED_CA_FILE="/home/work/_app/k8s/etcd/ssl/ca.pem"
ETCD_CLIENT_CERT_AUTH="true"
ETCD_PEER_CERT_FILE="/home/work/_app/k8s/etcd/ssl/server.pem"
ETCD_PEER_KEY_FILE="/home/work/_app/k8s/etcd/ssl/server-key.pem"
ETCD_PEER_TRUSTED_CA_FILE="/home/work/_app/k8s/etcd/ssl/ca.pem"
ETCD_PEER_CLIENT_CERT_AUTH="true"

EOF

7.5启动ETCD服务

[root@k8s-master01 ~]# ansible all -m raw -a "systemctl daemon-reload && systemctl enable etcd && systemctl start etcd && systemctl status etcd"

7.6检查ETCD服务运行状态

[root@k8s-master01 ~]# /home/work/_app/k8s/etcd/bin/etcdctl --ca-file=/home/work/_app/k8s/etcd/ssl/ca.pem --cert-file=/home/work/_app/k8s/etcd/ssl/server.pem --key-file=/home/work/_app/k8s/etcd/ssl/server-key.pem cluster-health
member 3b17aaa147134dd is healthy: got healthy result from https://10.0.0.10:2379
member 55fcbe0adaa45350 is healthy: got healthy result from https://10.0.0.13:2379
member cebdf10928a06f3c is healthy: got healthy result from https://10.0.0.11:2379
member f7a9c20602b8532e is healthy: got healthy result from https://10.0.0.12:2379
cluster is healthy

7.7检查ETCD集群成员信息

[root@k8s-master01 ~]# /home/work/_app/k8s/etcd/bin/etcdctl --ca-file=/home/work/_app/k8s/etcd/ssl/ca.pem --cert-file=/home/work/_app/k8s/etcd/ssl/server.pem --key-file=/home/work/_app/k8s/etcd/ssl/server-key.pem  member list
3b17aaa147134dd: name=etcd00 peerURLs=https://10.0.0.10:2380 clientURLs=https://10.0.0.10:2379 isLeader=false
55fcbe0adaa45350: name=etcd03 peerURLs=https://10.0.0.13:2380 clientURLs=https://10.0.0.13:2379 isLeader=true
cebdf10928a06f3c: name=etcd01 peerURLs=https://10.0.0.11:2380 clientURLs=https://10.0.0.11:2379 isLeader=false
f7a9c20602b8532e: name=etcd02 peerURLs=https://10.0.0.12:2380 clientURLs=https://10.0.0.12:2379 isLeader=false

8安装Flannel

8.1Flanneld 网络安装

Flannel 实质上是一种“覆盖网络(overlay network)”，也就是将TCP数据包装在另一种网络包里面进行路由转发和通信，目前已经支持UDP、VxLAN、AWS VPC和GCE路由等数据转发方式。Flannel 在 Kubernetes中用于配置第三层（网络层）网络结构。

Flannel 负责在集群中的多个节点之间提供第 3 层 IPv4 网络。Flannel 不控制容器如何与主机联网，只负责主机之间如何传输流量。但是，Flannel 确实为 Kubernetes 提供了 CNI 插件，并提供了与 Docker 集成的指导。

[^]: 没有 Flanneld 网络，Node节点间的 pod 不能通信，只能 Node 内通信。 有 Flanneld 服务启动时主要做了以下几步的工作： 从 ETCD 中获取 NetWork 的配置信息划分 Subnet，并在 ETCD 中进行注册,将子网信息记录到 /run/flannel/subnet.env 中

8.2向ETCD集群写入网段信息

[root@k8s-master01 _src]# /home/work/_app/k8s/etcd/bin/etcdctl --ca-file=/home/work/_app/k8s/etcd/ssl/ca.pem --cert-file=/home/work/_app/k8s/etcd/ssl/server.pem --key-file=/home/work/_app/k8s/etcd/ssl/server-key.pem --endpoints="https://10.0.0.10:2379,https://10.0.0.11:2379,https://10.0.0.12:2379,https://10.0.0.13:2379"  set /coreos.com/network/config  '{ "Network": "10.244.0.0/16", "Backend": {"Type": "vxlan"}}'

Flanneld 当前版本 (v0.11.0) 不支持 ETCD v3，所以使用 ETCD v2 API 写入配置 key 和网段数据； 写入的 Pod 网段 ${CLUSTER_CIDR} 必须是 /16 段地址，必须与 kube-controller-manager 的 –cluster-cidr 参数值一致；

8.3安装Flannel

[root@k8s-master01 _src]# ansible all -m shell -a "mkdir /home/work/_app/k8s/kubernetes/{bin,cfg,ssl} -p"
[root@k8s-master01 _src]# wget https://github.com/coreos/flannel/releases/download/v0.11.0/flannel-v0.11.0-linux-amd64.tar.gz
[root@k8s-master01 _src]# tar -xvf flannel-v0.11.0-linux-amd64.tar.gz
[root@k8s-master01 _src]# mv flanneld mk-docker-opts.sh /home/work/_app/k8s/kubernetes/bin/

mk-docker-opts.sh 脚本将分配给 Flanneld 的 Pod 子网网段信息写入 /run/flannel/docker 文件，后续 Docker 启动时 使用这个文件中的环境变量配置 docker0 网桥. Flanneld 使用系统缺省路由所在的接口与其它节点通信，对于有多个网络接口（如内网和公网）的节点，可以用 -iface 参数指定通信接口;

8.4配置flannel

创建 /home/work/_app/k8s/kubernetes/cfg/flanneld 文件并保存，写入以下内容：

[root@k8s-master01 _src]# cat << EOF | tee /home/work/_app/k8s/kubernetes/cfg/flanneld
FLANNEL_OPTIONS="--etcd-endpoints=https://10.0.0.10:2379,https://10.0.0.11:2379,https://10.0.0.12:2379,https://10.0.0.13:2379 -etcd-cafile=/home/work/_app/k8s/etcd/ssl/ca.pem -etcd-certfile=/home/work/_app/k8s/etcd/ssl/server.pem -etcd-keyfile=/home/work/_app/k8s/etcd/ssl/server-key.pem"
EOF

8.5创建Flannel系统启动文件

创建 /usr/lib/systemd/system/flanneld.service 文件并保存，内容如下：

[root@k8s-master01 _src]# cat << EOF | tee /usr/lib/systemd/system/flanneld.service
[Unit]
Description=Flanneld overlay address etcd agent
After=network-online.target network.target
Before=docker.service

[Service]
Type=notify
EnvironmentFile=/home/work/_app/k8s/kubernetes/cfg/flanneld
ExecStart=/home/work/_app/k8s/kubernetes/bin/flanneld --ip-masq $FLANNEL_OPTIONS
ExecStartPost=/home/work/_app/k8s/kubernetes/bin/mk-docker-opts.sh -k DOCKER_NETWORK_OPTIONS -d /run/flannel/subnet.env
Restart=on-failure

[Install]
WantedBy=multi-user.target
EOF

 

8.6配置 Docker 启动指定子网段

编辑 /usr/lib/systemd/system/docker.service 文件，内容如下：

[root@k8s-master01 _src]# cat << EOF | tee /usr/lib/systemd/system/docker.service
[Unit]
Description=Docker Application Container Engine
Documentation=https://docs.docker.com
After=network-online.target firewalld.service
Wants=network-online.target

[Service]
Type=notify
# 加入环境变量的配件文件，并在 ExecStart 附加参数
EnvironmentFile=/run/flannel/subnet.env
ExecStart=/usr/bin/dockerd $DOCKER_NETWORK_OPTIONS
ExecReload=/bin/kill -s HUP $MAINPID
LimitNOFILE=infinity
LimitNPROC=infinity
LimitCORE=infinity
TimeoutStartSec=0
Delegate=yes
KillMode=process
Restart=on-failure
StartLimitBurst=3
StartLimitInterval=60s

[Install]
WantedBy=multi-user.target
EOF

8.7分发flannel相关文件到其他节点

[root@k8s-master01 _src]# ansible other -m copy -a "src=/home/work/_app/k8s/kubernetes dest=/home/work/_app/k8s/ mode=0744"
[root@k8s-master01 _src]# ansible other  -a "ls /home/work/_app/k8s/kubernetes"
[root@k8s-master01 _src]# ansible other -m copy -a "src=/usr/lib/systemd/system/flanneld.service dest=/usr/lib/systemd/system/"
[root@k8s-master01 _src]# ansible other  -a "ls /usr/lib/systemd/system/flanneld.service"
[root@k8s-master01 _src]# ansible other -m copy -a "src=/usr/lib/systemd/system/docker.service dest=/usr/lib/systemd/system/"
[root@k8s-master01 _src]# ansible other  -a "ls /usr/lib/systemd/system/docker.service"

8.8启动flannel服务

ansible all -m shell -a "systemctl daemon-reload && systemctl stop docker && systemctl enable flanneld && systemctl start flanneld && systemctl start docker && systemctl status flanneld && systemctl status docker"

注意： <!--启动 Flannel 前要关闭 Docker 及相关的 kubelet 这样 Flannel 才会覆盖 docker0 网桥-->

8.9验证flannel

[root@k8s-master01 _src]# ansible all -m raw -a "cat /run/flannel/subnet.env && ip a|egrep -A 2 flannel"

查看网段是否符合预期

9.安装kubernetes

9.1生成 Kubernetes 证书请求的JSON请求文件

[root@k8s-master01 ~]# cd /home/work/_app/k8s/kubernetes/ssl/
[root@k8s-master01 ssl]# cat << EOF | tee ca-config.json
{
  "signing": {
    "default": {
      "expiry": "8760h"
    },
    "profiles": {
      "server": {
        "usages": [
          "signing",
          "key encipherment",
          "server auth"
        ],
        "expiry": "8760h"
      },
      "client": {
        "usages": [
          "signing",
          "key encipherment",
          "client auth"
        ],
        "expiry": "8760h"
      }
    }
  }
}
EOF
#生成 ca-config.json

9.2生成 Kubernetes CA 配置文件和证书

[root@k8s-master01 ssl]# cat << EOF | tee ca-csr.json
{
    "CN": "kubernetes CA",
    "key": {
        "algo": "rsa",
        "size": 2048
    },
    "names": [
        {
            "C": "CN",
            "L": "Beijing",
            "ST": "Beijing",
            "O": "k8s",
            "OU": "System"
        }
    ]
}
EOF
#生成 ca-csr.json

初始化一个 Kubernetes CA 证书

[root@k8s-master01 ssl]# ls
ca-config.json  ca-csr.json
[root@k8s-master01 ssl]# cfssl gencert -initca ca-csr.json | cfssljson -bare ca
2019/11/03 17:31:20 [INFO] generating a new CA key and certificate from CSR
2019/11/03 17:31:20 [INFO] generate received request
2019/11/03 17:31:20 [INFO] received CSR
2019/11/03 17:31:20 [INFO] generating key: rsa-2048
2019/11/03 17:31:21 [INFO] encoded CSR
2019/11/03 17:31:21 [INFO] signed certificate with serial number 363442360986653592971594241252855429808195119806
[root@k8s-master01 ssl]# ls
ca-config.json  ca.csr  ca-csr.json  ca-key.pem  ca.pem
#生成 ca.csr ca-key.pem  ca.pem

9.3生成 Kube API Server 配置文件和证书

[root@k8s-master01 ssl]# cat << EOF | tee kube-apiserver-server-csr.json
{
    "CN": "kubernetes",
    "hosts": [
      "127.0.0.1",
      "10.0.0.2",  
      "10.0.0.10", 
      "kubernetes",
      "kubernetes.default",
      "kubernetes.default.svc",
      "kubernetes.default.svc.cluster",
      "kubernetes.default.svc.cluster.local"
    ],
    "key": {
        "algo": "rsa",
        "size": 2048
    },
    "names": [
        {
            "C": "CN",
            "L": "Beijing",
            "ST": "Beijing",
            "O": "k8s",
            "OU": "API Server"
        }
    ]
}
EOF
# 生成 kube-apiserver-server-csr.json

生成 kube-apiserver 证书

[root@k8s-master01 ssl]# ls
ca-config.json  ca.csr  ca-csr.json  ca-key.pem  ca.pem  kube-apiserver-server-csr.json
[root@k8s-master01 ssl]# cfssl gencert -ca=ca.pem -ca-key=ca-key.pem -config=ca-config.json -profile=server kube-apiserver-server-csr.json | cfssljson -bare kube-apiserver-server
2019/11/03 17:41:39 [INFO] generate received request
2019/11/03 17:41:39 [INFO] received CSR
2019/11/03 17:41:39 [INFO] generating key: rsa-2048
2019/11/03 17:41:39 [INFO] encoded CSR
2019/11/03 17:41:39 [INFO] signed certificate with serial number 40258078055292579176801571476092842140605641659
2019/11/03 17:41:39 [WARNING] This certificate lacks a "hosts" field. This makes it unsuitable for
websites. For more information see the Baseline Requirements for the Issuance and Management
of Publicly-Trusted Certificates, v.1.1.6, from the CA/Browser Forum (https://cabforum.org);
specifically, section 10.2.3 ("Information Requirements").
[root@k8s-master01 ssl]# ls
ca-config.json  ca.csr  ca-csr.json  ca-key.pem  ca.pem  kube-apiserver-server.csr  kube-apiserver-server-csr.json  kube-apiserver-server-key.pem  kube-apiserver-server.pem
# 生成kube-apiserver-server.csr kube-apiserver-server-key.pem  kube-apiserver-server.pem

9.4生成 kubelet client 配置文件和证书

创建证书配置文件

[root@k8s-master01 ssl]#  cat << EOF | tee kubelet-client-csr.json
{
  "CN": "kubelet",
  "hosts": [""],
  "key": {
    "algo": "rsa",
    "size": 2048
  },
  "names": [
    {
      "C": "CN",
      "L": "Beijing",
      "O": "k8s",
      "OU": "Kubelet",
      "ST": "Beijing"
    }
  ]
}
EOF

生成 kubelet client证书

[root@k8s-master01 ssl]# cfssl gencert -ca=ca.pem -ca-key=ca-key.pem -config=ca-config.json -profile=client kubelet-client-csr.json | cfssljson -bare kubelet-client
2019/11/03 17:54:22 [INFO] generate received request
2019/11/03 17:54:22 [INFO] received CSR
2019/11/03 17:54:22 [INFO] generating key: rsa-2048
2019/11/03 17:54:22 [INFO] encoded CSR
2019/11/03 17:54:22 [INFO] signed certificate with serial number 584708907544786851775012737222915080474397058953
2019/11/03 17:54:22 [WARNING] This certificate lacks a "hosts" field. This makes it unsuitable for
websites. For more information see the Baseline Requirements for the Issuance and Management
of Publicly-Trusted Certificates, v.1.1.6, from the CA/Browser Forum (https://cabforum.org);
specifically, section 10.2.3 ("Information Requirements").
[root@k8s-master01 ssl]# ls
ca-config.json  ca-csr.json  ca.pem                     kube-apiserver-server-csr.json  kube-apiserver-server.pem  kubelet-client-csr.json  kubelet-client.pem
ca.csr          ca-key.pem   kube-apiserver-server.csr  kube-apiserver-server-key.pem   kubelet-client.csr         kubelet-client-key.pem
#生成 kube-apiserver-server-key.pem   kubelet-client.csr         kubelet-client-key.pem

9.5生成 Kube-Proxy 配置文件和证书

  创建证书配置文件

[root@k8s-master01 ssl]# cat << EOF | tee kube-proxy-client-csr.json
{
  "CN": "system:kube-proxy",
  "hosts": [""],
  "key": {
    "algo": "rsa",
    "size": 2048
  },
  "names": [
    {
      "C": "CN",
      "L": "Beijing",
      "O": "k8s",
      "OU": "System",
      "ST": "Beijing"
    }
  ]
}
EOF
# 生成 kube-proxy-client-csr.json

 生成 Kube-Proxy 证书

[root@k8s-master01 ssl]# cfssl gencert -ca=ca.pem -ca-key=ca-key.pem -config=ca-config.json -profile=client kube-proxy-client-csr.json | cfssljson -bare kube-proxy-client
2019/11/04 11:28:01 [INFO] generate received request
2019/11/04 11:28:01 [INFO] received CSR
2019/11/04 11:28:01 [INFO] generating key: rsa-2048
2019/11/04 11:28:01 [INFO] encoded CSR
2019/11/04 11:28:01 [INFO] signed certificate with serial number 67285254948566587248810360676603036529501371119
2019/11/04 11:28:01 [WARNING] This certificate lacks a "hosts" field. This makes it unsuitable for
websites. For more information see the Baseline Requirements for the Issuance and Management
of Publicly-Trusted Certificates, v.1.1.6, from the CA/Browser Forum (https://cabforum.org);
specifically, section 10.2.3 ("Information Requirements").
[root@k8s-master01 ssl]# ls
ca-config.json  ca-csr.json  ca.pem                     kube-apiserver-server-csr.json  kube-apiserver-server.pem  kubelet-client-csr.json  kubelet-client.pem     kube-proxy-client-csr.json  kube-proxy-client.pem
ca.csr          ca-key.pem   kube-apiserver-server.csr  kube-apiserver-server-key.pem   kubelet-client.csr         kubelet-client-key.pem   kube-proxy-client.csr  kube-proxy-client-key.pem
#生成   kubelet-client-key.pem   kube-proxy-client.csr  kube-proxy-client-key.pem

9.6生成 kubectl 管理员配置文件和证书

  创建 kubectl 管理员证书配置文件

[root@k8s-master01 ssl]# cat << EOF | tee kubernetes-admin-user.csr.json
{
  "CN": "admin",
  "hosts": [""],
  "key": {
    "algo": "rsa",
    "size": 2048
  },
  "names": [
    {
      "C": "CN",
      "L": "Beijing",
      "O": "k8s",
      "OU": "Cluster Admins",
      "ST": "Beijing"
    }
  ]
}
EOF

生成 kubectl 管理员证书

[root@k8s-master01 ssl]# cfssl gencert -ca=ca.pem -ca-key=ca-key.pem -config=ca-config.json -profile=client kubernetes-admin-user.csr.json | cfssljson -bare kubernetes-admin-user
2019/11/04 11:30:17 [INFO] generate received request
2019/11/04 11:30:17 [INFO] received CSR
2019/11/04 11:30:17 [INFO] generating key: rsa-2048
2019/11/04 11:30:17 [INFO] encoded CSR
2019/11/04 11:30:17 [INFO] signed certificate with serial number 177011301186637544930154578732184051651945219290
2019/11/04 11:30:17 [WARNING] This certificate lacks a "hosts" field. This makes it unsuitable for
websites. For more information see the Baseline Requirements for the Issuance and Management
of Publicly-Trusted Certificates, v.1.1.6, from the CA/Browser Forum (https://cabforum.org);
specifically, section 10.2.3 ("Information Requirements").
[root@k8s-master01 ssl]# ls
ca-config.json  ca-key.pem                 kube-apiserver-server-csr.json  kubelet-client.csr       kubelet-client.pem          kube-proxy-client-key.pem  kubernetes-admin-user.csr.json
ca.csr          ca.pem                     kube-apiserver-server-key.pem   kubelet-client-csr.json  kube-proxy-client.csr       kube-proxy-client.pem      kubernetes-admin-user-key.pem
ca-csr.json     kube-apiserver-server.csr  kube-apiserver-server.pem       kubelet-client-key.pem   kube-proxy-client-csr.json  kubernetes-admin-user.csr  kubernetes-admin-user.pem 
#生成 kubernetes-admin-user.csr kubernetes-admin-user-key.pem kubernetes-admin-user.pem

9.7将相关证书复制到 Kubernetes Node 节点

[root@k8s-master01 ssl]# ansible other -m copy -a "src=/home/work/_app/k8s/kubernetes/ssl/ dest=/home/work/_app/k8s/kubernetes/ssl/"
[root@k8s-master01 ssl]# ansible all -m shell -a "ls -lrt /home/work/_app/k8s/kubernetes/ssl/"

10.部署kubernetes master

 Kubernetes Master 节点运行如下组件：

• APIServer APIServer负责对外提供RESTful的kubernetes API的服务，它是系统管理指令的统一接口，任何对资源的增删该查都要交给APIServer处理后再交给etcd，如图，kubectl(kubernetes提供的客户端工具，该工具内部是对kubernetes API的调用）是直接和APIServer交互的。

• Schedule schedule负责调度Pod到合适的Node上，如果把scheduler看成一个黑匣子，那么它的输入是pod和由多个Node组成的列表，输出是Pod和一个Node的绑定。 kubernetes目前提供了调度算法，同样也保留了接口。用户根据自己的需求定义自己的调度算法。

• Controller manager 如果APIServer做的是前台的工作的话，那么controller manager就是负责后台的。每一个资源都对应一个控制器。而control manager就是负责管理这些控制器的，比如我们通过APIServer创建了一个Pod，当这个Pod创建成功后，APIServer的任务就算完成了。

• ETCD etcd是一个高可用的键值存储系统，kubernetes使用它来存储各个资源的状态，从而实现了Restful的API。

• Flannel 默认没有flanneld网络，Node节点间的pod不能通信，只能Node内通信，Flannel从etcd中获取network的配置信息 划分subnet，并在etcd中进行注册 将子网信息记录

kube-scheduler 和 kube-controller-manager 可以以集群模式运行，通过 leader 选举产生一个工作进程，其它进程处于阻塞模式。

10.1、下载文件并安装 Kubernetes Server

[root@k8s-master01 ssl]# cd /home/work/_src/
[root@k8s-master01 _src]# wget https://dl.k8s.io/v1.13.0/kubernetes-server-linux-amd64.tar.gz
[root@k8s-master01 _src]# tar -xzvf kubernetes-server-linux-amd64.tar.gz
[root@k8s-master01 _src]# ansible all -m shell -a "mkdir /home/work/_src/kubernetes/server/bin/ -p"
[root@k8s-master01 bin]# cp kube-scheduler kube-apiserver kube-controller-manager kubectl kubelet kube-proxy /home/work/_app/k8s/kubernetes/bin/

10.2 分发文件到其他节点

kubelet kube-proxy kubectl

[root@k8s-master01 bin]#  ansible other -m copy  -a "src=/home/work/_app/k8s/kubernetes/bin/kubelet  dest=/home/work/_app/k8s/kubernetes/bin/ mode=0744"
[root@k8s-master01 bin]#  ansible other -m copy  -a "src=/home/work/_app/k8s/kubernetes/bin/kube-proxy  dest=/home/work/_app/k8s/kubernetes/bin/ mode=0744"
[root@k8s-master01 bin]#  ansible other -m copy  -a "src=/home/work/_app/k8s/kubernetes/bin/kubectl  dest=/home/work/_app/k8s/kubernetes/bin/ mode=0744"

10.3部署apiserver

创建 TLS Bootstrapping Token

[root@k8s-master01 bin]# head -c 16 /dev/urandom | od -An -t x | tr -d ' '
f03e0561ac6ff77cdd835693f557788d

这里我们生成的随机Token是f03e0561ac6ff77cdd835693f557788d，记下来后面要用到

[root@k8s-master01 bin]# cat /home/work/_app/k8s/kubernetes/cfg/token-auth-file
f03e0561ac6ff77cdd835693f557788d,kubelet-bootstrap,10001,"system:kubelet-bootstrap"

10.3.1创建 Apiserver 配置文件

  创建 /home/work/_app/k8s/kubernetes/cfg/kube-apiserver 文件并保存，内容如下：

[root@k8s-master01 bin]# cat /home/work/_app/k8s/kubernetes/cfg/kube-apiserver
KUBE_APISERVER_OPTS="--logtostderr=true \
--v=4 \
--etcd-servers=https://10.0.0.10:2379,https://10.0.0.11:2379,https://10.0.0.12:2379,https://10.0.0.13:2379 \
--bind-address=10.0.0.10 \
--secure-port=6443 \
--advertise-address=10.0.0.10 \
--allow-privileged=true \
--service-cluster-ip-range=10.244.0.0/16 \
--enable-admission-plugins=NamespaceLifecycle,LimitRanger,SecurityContextDeny,ServiceAccount,ResourceQuota,NodeRestriction \
--authorization-mode=RBAC,Node \
--enable-bootstrap-token-auth \
--token-auth-file=/home/work/_app/k8s/kubernetes/cfg/token-auth-file \
--service-node-port-range=30000-50000 \
--tls-cert-file=/home/work/_app/k8s/kubernetes/ssl/kube-apiserver-server.pem  \
--tls-private-key-file=/home/work/_app/k8s/kubernetes/ssl/kube-apiserver-server-key.pem \
--client-ca-file=/home/work/_app/k8s/kubernetes/ssl/ca.pem \
--service-account-key-file=/home/work/_app/k8s/kubernetes/ssl/ca-key.pem \
--etcd-cafile=/home/work/_app/k8s/etcd/ssl/ca.pem \
--etcd-certfile=/home/work/_app/k8s/etcd/ssl/server.pem \
--etcd-keyfile=/home/work/_app/k8s/etcd/ssl/server-key.pem"

10.3.2创建 Apiserver 启动文件

创建 /usr/lib/systemd/system/kube-apiserver.service 文件并保存，内容如下:

[root@k8s-master01 bin]# cat /usr/lib/systemd/system/kube-apiserver.service
[Unit]
Description=Kubernetes API Server
Documentation=https://github.com/kubernetes/kubernetes

[Service]
EnvironmentFile=-/home/work/_app/k8s/kubernetes/cfg/kube-apiserver
ExecStart=/home/work/_app/k8s/kubernetes/bin/kube-apiserver $KUBE_APISERVER_OPTS
Restart=on-failure

[Install]
WantedBy=multi-user.target

10.2.3启动 Kube Apiserver 服务

[root@k8s-master01 bin]# systemctl daemon-reload && systemctl enable kube-apiserver && systemctl start kube-apiserver && systemctl status kube-apiserver

 

10.4部署 Scheduler

创建 /home/work/_app/k8s/kubernetes/cfg/kube-scheduler 文件并保存，内容如下：

[root@k8s-master01 bin]#  cat /home/work/_app/k8s/kubernetes/cfg/kube-scheduler
KUBE_SCHEDULER_OPTS="--logtostderr=true --v=4 --master=127.0.0.1:8080 --leader-elect"

10.4.1创建 Kube-scheduler 系统启动文件

创建 /usr/lib/systemd/system/kube-scheduler.service 文件并保存，内容如下：

[root@k8s-master01 bin]#  cat /usr/lib/systemd/system/kube-scheduler.service
[Unit]
Description=Kubernetes Scheduler
Documentation=https://github.com/kubernetes/kubernetes

[Service]
EnvironmentFile=-/home/work/_app/k8s/kubernetes/cfg/kube-scheduler
ExecStart=/home/work/_app/k8s/kubernetes/bin/kube-scheduler $KUBE_SCHEDULER_OPTS
Restart=on-failure

[Install]
WantedBy=multi-user.target

10.4.2启动 Kube-scheduler 服务

[root@k8s-master01 bin]# systemctl daemon-reload && systemctl enable kube-scheduler && systemctl start kube-scheduler && systemctl status kube-scheduler

10.5部署 Kube-Controller-Manager 组件

10.5.1创建 kube-controller-manager 配置文件

 创建 /home/work/_app/k8s/kubernetes/cfg/kube-controller-manager 文件并保存，内容如下：

[root@k8s-master01 bin]# cat /home/work/_app/k8s/kubernetes/cfg/kube-controller-manager
KUBE_CONTROLLER_MANAGER_OPTS="--logtostderr=true \
--v=4 \
--master=127.0.0.1:8080 \
--leader-elect=true \
--address=127.0.0.1 \
--service-cluster-ip-range=10.244.0.0/16 \
--cluster-name=kubernetes \
--cluster-signing-cert-file=/home/work/_app/k8s/kubernetes/ssl/ca.pem \
--cluster-signing-key-file=/home/work/_app/k8s/kubernetes/ssl/ca-key.pem  \
--root-ca-file=/home/work/_app/k8s/kubernetes/ssl/ca.pem \
--service-account-private-key-file=/home/work/_app/k8s/kubernetes/ssl/ca-key.pem"

10.5.2创建 kube-controller-manager 系统启动文件

创建 /usr/lib/systemd/system/kube-controller-manager.service 文件并保存，内容如下

[root@k8s-master01 bin]# cat /usr/lib/systemd/system/kube-controller-manager.service
[Unit]
Description=Kubernetes Controller Manager
Documentation=https://github.com/kubernetes/kubernetes

[Service]
EnvironmentFile=-/home/work/_app/k8s/kubernetes/cfg/kube-controller-manager
ExecStart=/home/work/_app/k8s/kubernetes/bin/kube-controller-manager $KUBE_CONTROLLER_MANAGER_OPTS
Restart=on-failure

[Install]
WantedBy=multi-user.target

10.5.3启动 kube-controller-manager 服务

[root@k8s-master01 bin]# systemctl daemon-reload && systemctl enable kube-controller-manager && systemctl start kube-controller-manager && systemctl status kube-controller-manager

10.6验证API-Server服务

将 kubectl 加入到$PATH变量中

[root@k8s-master01 bin]# echo "PATH=/home/work/_app/k8s/kubernetes/bin:$PATH:$HOME/bin" >> /etc/profile
[root@k8s-master01 bin]# source /etc/profile

查看节点状态

[root@k8s-master01 ~]# kubectl get cs,nodes
NAME                                 STATUS    MESSAGE             ERROR
componentstatus/etcd-2               Healthy   {"health":"true"}   
componentstatus/controller-manager   Healthy   ok                  
componentstatus/scheduler            Healthy   ok                  
componentstatus/etcd-1               Healthy   {"health":"true"}   
componentstatus/etcd-3               Healthy   {"health":"true"}   
componentstatus/etcd-0               Healthy   {"health":"true"}   

10.7部署kubelet

10.7.1创建 bootstrap.kubeconfig、kube-proxy.kubeconfig 配置文件

创建 /home/work/_app/k8s/kubernetes/cfg/env.sh 文件并保存，内容如下：

[root@k8s-master01 ~]# cd /home/work/_app/k8s/kubernetes/cfg/
[root@k8s-master01 ~]# cat  env.sh 
#!/bin/bash
#创建kubelet bootstrapping kubeconfig 
BOOTSTRAP_TOKEN=f03e0561ac6ff77cdd835693f557788d
KUBE_APISERVER="https://10.0.0.10:6443"
#设置集群参数
kubectl config set-cluster kubernetes \
  --certificate-authority=/home/work/_app/k8s/kubernetes/ssl/ca.pem \
  --embed-certs=true \
  --server=${KUBE_APISERVER} \
  --kubeconfig=bootstrap.kubeconfig
 
#设置客户端认证参数
kubectl config set-credentials kubelet-bootstrap \
  --token=${BOOTSTRAP_TOKEN} \
  --kubeconfig=bootstrap.kubeconfig
 
# 设置上下文参数
kubectl config set-context default \
  --cluster=kubernetes \
  --user=kubelet-bootstrap \
  --kubeconfig=bootstrap.kubeconfig
 
# 设置默认上下文
kubectl config use-context default --kubeconfig=bootstrap.kubeconfig
 
#----------------------
 
# 创建kube-proxy kubeconfig文件
 
kubectl config set-cluster kubernetes \
  --certificate-authority=/home/work/_app/k8s/kubernetes/ssl/ca.pem \
  --embed-certs=true \
  --server=${KUBE_APISERVER} \
  --kubeconfig=kube-proxy.kubeconfig
 
kubectl config set-credentials kube-proxy \
  --client-certificate=/home/work/_app/k8s/kubernetes/ssl/kube-proxy-client.pem \
  --client-key=/home/work/_app/k8s/kubernetes/ssl/kube-proxy-client-key.pem \
  --embed-certs=true \
  --kubeconfig=kube-proxy.kubeconfig
 
kubectl config set-context default \
  --cluster=kubernetes \
  --user=kube-proxy \
  --kubeconfig=kube-proxy.kubeconfig
 
kubectl config use-context default --kubeconfig=kube-proxy.kubeconfig

BOOTSTRAP_TOKEN`使用在创建 `TLS Bootstrapping Token` 生成的`f03e0561ac6ff77cdd835693f557788d`

  执行脚本:

[root@k8s-master01 cfg]# sh env.sh 
Cluster "kubernetes" set.
User "kubelet-bootstrap" set.
Context "default" created.
Switched to context "default".
Cluster "kubernetes" set.
User "kube-proxy" set.
Context "default" created.
Switched to context "default".
[root@k8s-master01 cfg]# ls
bootstrap.kubeconfig  env.sh  flanneld  kube-apiserver  kube-controller-manager  kube-proxy.kubeconfig  kube-scheduler  token-auth-file
# 执行脚本生成 bootstrap.kubeconfig   kube-proxy.kubeconfig

  将 bootstrap.kubeconfig、kube-proxy.kubeconfig 复制到其他节点

[root@k8s-master01 cfg]# ansible other -m copy -a "src=/home/work/_app/k8s/kubernetes/cfg/kube-proxy.kubeconfig dest=/home/work/_app/k8s/kubernetes/cfg/"
[root@k8s-master01 cfg]# ansible other -m copy -a "src=/home/work/_app/k8s/kubernetes/cfg/bootstrap.kubeconfig  dest=/home/work/_app/k8s/kubernetes/cfg/"
[root@k8s-master01 cfg]# ansible other -m shell -a "ls /home/work/_app/k8s/kubernetes/cfg/bootstrap.kubeconfig"
[root@k8s-master01 cfg]# ansible other -m shell -a "ls /home/work/_app/k8s/kubernetes/cfg/kube-proxy.kubeconfig"

10.7.2创建 kubelet 配置文件

创建 /home/work/_app/k8s/kubernetes/cfg/kubelet.config 参数配置文件并保存，内容如下：

[root@k8s-master01 cfg]# cat /home/work/_app/k8s/kubernetes/cfg/kubelet.config
kind: KubeletConfiguration
apiVersion: kubelet.config.k8s.io/v1beta1
address: 10.0.0.10
port: 10250
readOnlyPort: 10255
cgroupDriver: cgroupfs
clusterDNS: ["10.244.0.1"]
clusterDomain: cluster.local.
failSwapOn: false
authentication:
  anonymous:
    enabled: true

创建 /home/work/_app/k8s/kubernetes/cfg/kubelet 启动参数文件并保存，内容如下：

[root@k8s-master01 cfg]# cat /home/work/_app/k8s/kubernetes/cfg/kubelet
KUBELET_OPTS="--logtostderr=true \
--v=4 \
--hostname-override=10.0.0.10 \
--kubeconfig=/home/work/_app/k8s/kubernetes/cfg/kubelet.kubeconfig \
--bootstrap-kubeconfig=/home/work/_app/k8s/kubernetes/cfg/bootstrap.kubeconfig \
--config=/home/work/_app/k8s/kubernetes/cfg/kubelet.config \
--cert-dir=/home/work/_app/k8s/kubernetes/ssl_cert \
--pod-infra-container-image=registry.cn-hangzhou.aliyuncs.com/google-containers/pause-amd64:3.0"

当 kubelet 启动时，如果通过 --kubeconfig 指定的文件不存在，则通过 --bootstrap-kubeconfig 指定的 bootstrap kubeconfig 用于从API服务器请求客户端证书。 在通过 kubelet 批准证书请求时，引用生成的密钥和证书将放在 --cert-dir 目录中。

  

10.7.3、将 kubelet-bootstrap 用户绑定到系统集群角色

[root@k8s-master01 cfg]# kubectl create clusterrolebinding kubelet-bootstrap --clusterrole=system:node-bootstrapper --user=kubelet-bootstrap
clusterrolebinding.rbac.authorization.k8s.io/kubelet-bootstrap created

10.7.4、创建 kubelet 系统启动文件

  创建 /usr/lib/systemd/system/kubelet.service 并保存，内容如下：

[root@k8s-master01 cfg]# cat /usr/lib/systemd/system/kubelet.service
[Unit]
Description=Kubernetes Kubelet
After=docker.service
Requires=docker.service
 
[Service]
EnvironmentFile=/home/work/_app/k8s/kubernetes/cfg/kubelet
ExecStart=/home/work/_app/k8s/kubernetes/bin/kubelet $KUBELET_OPTS
Restart=on-failure
KillMode=process
 
[Install]
WantedBy=multi-user.target

10.7.5启动kubelet服务

[root@k8s-master01 cfg]# systemctl daemon-reload && systemctl enable kubelet && systemctl start kubelet &&systemctl status kubelet

10.8批准 Master 加入集群

CSR 可以在内置批准流程之外做手动批准加入集群。 管理员也可以使用 kubectl 手动批准证书请求。 管理员可以使用 kubectl get csr 列出 CSR 请求， 并使用 kubectl describe csr 列出详细描述。 管理员也可以使用 kubectl certificate approve 或 kubectl certificate deny 工具批准或拒绝 CSR 请求。

 

10.8.1查看 CSR 列表

[root@k8s-master01 cfg]# kubectl get csr
NAME                                                   AGE     REQUESTOR        CONDITION
node-csr-JxDAOeYmXgaOHWVraQT7ScZ2zt0mj52cC9vA-4WZMUA   2m11s   kubelet-bootstrap  Pending

10.8.2审批加入集群

[root@k8s-master01 cfg]# kubectl certificate approve node-csr-JxDAOeYmXgaOHWVraQT7ScZ2zt0mj52cC9vA-4WZMUA
certificatesigningrequest.certificates.k8s.io/node-csr-JxDAOeYmXgaOHWVraQT7ScZ2zt0mj52cC9vA-4WZMUA approved

10.8.3验证master是否加入集群

 再次查看 CSR 列表

[root@k8s-master01 cfg]# kubectl get csr
NAME                                                   AGE     REQUESTOR           CONDITION
node-csr-JxDAOeYmXgaOHWVraQT7ScZ2zt0mj52cC9vA-4WZMUA   5m26s   kubelet-bootstrap   Approved,Issued

10.9部署 kube-proxy 组件

kube-proxy 运行在所有 Node 节点上，它监听 apiserver 中 service 和 Endpoint 的变化情况，创建路由规则来进行服务负载均衡，以下操作以 k8s-master01 为例

10.9.1创建 kube-proxy 参数配置文件

  创建 /home/work/_app/k8s/kubernetes/cfg/kube-proxy 配置文件并保存，内容如下：

[root@k8s-master01 cfg]# cat /home/work/_app/k8s/kubernetes/cfg/kube-proxy
KUBE_PROXY_OPTS="--logtostderr=true \
--v=4 \
--hostname-override=10.0.0.10 \
--cluster-cidr=10.244.0.0/16 \
--kubeconfig=/home/work/_app/k8s/kubernetes/cfg/kube-proxy.kubeconfig"

--hostname-override在不同的节点处，要换成节点的IP

10.9.2创建 kube-proxy 系统启动文件

  创建 /usr/lib/systemd/system/kube-proxy.service 文件并保存，内容如下：

[root@k8s-master01 cfg]# cat  /usr/lib/systemd/system/kube-proxy.service
[Unit]
Description=Kubernetes Proxy
After=network.target
 
[Service]
EnvironmentFile=-/home/work/_app/k8s/kubernetes/cfg/kube-proxy
ExecStart=/home/work/_app/k8s/kubernetes/bin/kube-proxy $KUBE_PROXY_OPTS
Restart=on-failure
 
[Install]
WantedBy=multi-user.target 

10.9.3启动kube-proxy

[root@k8s-master01 cfg]# systemctl daemon-reload && systemctl enable kube-proxy &&  systemctl start kube-proxy && systemctl status kube-proxy

11.验证 Server 服务

[root@k8s-master01 cfg]# kubectl get cs,nodes
NAME                                 STATUS    MESSAGE             ERROR
componentstatus/etcd-1               Healthy   {"health":"true"}   
componentstatus/etcd-3               Healthy   {"health":"true"}   
componentstatus/etcd-0               Healthy   {"health":"true"}   
componentstatus/scheduler            Healthy   ok                  
componentstatus/etcd-2               Healthy   {"health":"true"}   
componentstatus/controller-manager   Healthy   ok                  

NAME             STATUS   ROLES    AGE     VERSION
node/10.0.0.10   Ready    <none>   8m12s   v1.13.0

12.Kubernetes Node 节点加入集群

Kubernetes Node 节点运行如下组件：

• Proxy: 该模块实现了kubernetes中的服务发现和反向代理功能。kube-proxy支持TCP和UDP连接转发，默认基Round Robin算法将客户端流量转发到与service对应的一组后端pod。服务发现方面，kube-proxy使用etcd的watch机制监控集群中service和endpoint对象数据的动态变化，并且维护一个service到endpoint的映射关系，从而保证了后端pod的IP变化不会对访问者造成影响，另外，kube-proxy还支持session affinity。

• Kublet kublet是Master在每个Node节点上面的agent，是Node节点上面最重要的模块，它负责维护和管理该Node上的所有容器，但是如果容器不是通过kubernetes创建的，它并不会管理。本质上，它负责使Pod的运行状态与期望的状态一致。 kublet 启动时自动向 kube-apiserver 注册节点信息，内置的 cadvisor 统计和监控节点的资源使用情况; 为确保安全，只开启接收 https 请求的安全端口，对请求进行认证和授权，拒绝未授权的访问(如apiserver、heapster)

• Flannel 默认没有flanneld网络，Node节点间的pod不能通信，只能Node内通信，Flannel从etcd中获取network的配置信息 划分subnet，并在etcd中进行注册 将子网信息记录

• ETCD ETCD是一个高可用的键值存储系统，kubernetes使用它来存储各个资源的状态，从而实现了Restful的API。

12.1创建 kubelet 配置文件

在k8s-master02 :   创建 /home/work/_app/k8s/kubernetes/cfg/kubelet.config 参数配置文件并保存，内容如下：

[root@k8s-master02 ~]# cat /home/work/_app/k8s/kubernetes/cfg/kubelet.config
kind: KubeletConfiguration
apiVersion: kubelet.config.k8s.io/v1beta1
address: 10.0.0.11
port: 10250
readOnlyPort: 10255
cgroupDriver: cgroupfs
clusterDNS: ["10.244.0.1"]
clusterDomain: cluster.local.
failSwapOn: false
authentication:
  anonymous:
    enabled: true

在k8s-node01:   创建 /home/work/_app/k8s/kubernetes/cfg/kubelet.config 参数配置文件并保存，内容如下：

[root@k8s-node01 ~]# cat /home/work/_app/k8s/kubernetes/cfg/kubelet.config
kind: KubeletConfiguration
apiVersion: kubelet.config.k8s.io/v1beta1
address: 10.0.0.12
port: 10250
readOnlyPort: 10255
cgroupDriver: cgroupfs
clusterDNS: ["10.244.0.1"]
clusterDomain: cluster.local.
failSwapOn: false
authentication:
  anonymous:
    enabled: true

在k8s-node02 :   创建 /home/work/_app/k8s/kubernetes/cfg/kubelet.config 参数配置文件并保存，内容如下：

[root@k8s-node02 ~]# cat /home/work/_app/k8s/kubernetes/cfg/kubelet.config
kind: KubeletConfiguration
apiVersion: kubelet.config.k8s.io/v1beta1
address: 10.0.0.13
port: 10250
readOnlyPort: 10255
cgroupDriver: cgroupfs
clusterDNS: ["10.244.0.1"]
clusterDomain: cluster.local.
failSwapOn: false
authentication:
  anonymous:
    enabled: true

创建 /home/work/_app/k8s/kubernetes/cfg/kubelet 启动参数文件并保存，内容如下：

[root@k8s-master02 ~]# cat /home/work/_app/k8s/kubernetes/cfg/kubelet
KUBELET_OPTS="--logtostderr=true \
--v=4 \
--hostname-override=10.0.0.11 \
--kubeconfig=/home/work/_app/k8s/kubernetes/cfg/kubelet.kubeconfig \
--bootstrap-kubeconfig=/home/work/_app/k8s/kubernetes/cfg/bootstrap.kubeconfig \
--config=/home/work/_app/k8s/kubernetes/cfg/kubelet.config \
--cert-dir=/home/work/_app/k8s/kubernetes/ssl_cert \
--pod-infra-container-image=registry.cn-hangzhou.aliyuncs.com/google-containers/pause-amd64:3.0"

--hostname-override 这个参数分别替换成各自节点的ip

12.2kubelet启动文件

[root@k8s-master01 ~]# ansible other -m copy -a "src=/usr/lib/systemd/system/kubelet.service dest=/usr/lib/systemd/system/"

[root@k8s-master01 ~]# ansible other -m shell -a "ls /usr/lib/systemd/system/kubelet.service"

12.3分别启动kubelet

[root@k8s-master01 ~]#ansible all -m shell -a  "systemctl daemon-reload && systemctl enable kubelet && systemctl start kubelet && systemctl status kubelet" |grep running

12.4批准 Node 加入集群

  查看 CSR 列表，可以看到节点有 Pending 请求

[root@k8s-master01 ~]# kubectl get csr
NAME                                                   AGE    REQUESTOR           CONDITION
node-csr-JxDAOeYmXgaOHWVraQT7ScZ2zt0mj52cC9vA-4WZMUA   30m    kubelet-bootstrap   Approved,Issued
node-csr-MrvfF1bll1SSDFHHOs5bx4Xw3peQ3vz_mNYdeY6hrzA   114s   kubelet-bootstrap   Pending
node-csr-Tl7t4oF9mp_obWTf5F_55whMrcWStRc3f4TYJtKNTVU   114s   kubelet-bootstrap   Pending
node-csr-VSttFy1WE5JiBIsfUdKTuCi-EA_DdkLihE1csGCbcXE   114s   kubelet-bootstrap   Pending

通过以下命令，查看请求的详细信息，能够看到是 k8s-master02 的IP地址10.0.0.101发来的请求

[root@k8s-master01 ~]# kubectl describe csr node-csr-MrvfF1bll1SSDFHHOs5bx4Xw3peQ3vz_mNYdeY6hrzA
Name:               node-csr-MrvfF1bll1SSDFHHOs5bx4Xw3peQ3vz_mNYdeY6hrzA
Labels:             <none>
Annotations:        <none>
CreationTimestamp:  Mon, 04 Nov 2019 20:05:08 +0800
Requesting User:    kubelet-bootstrap
Status:             Pending
Subject:
         Common Name:    system:node:10.0.0.11
         Serial Number:  
         Organization:   system:nodes
Events:  <none>

审批pendding的节点加入集群

[root@k8s-master01 ~]# kubectl certificate approve node-csr-MrvfF1bll1SSDFHHOs5bx4Xw3peQ3vz_mNYdeY6hrzA
certificatesigningrequest.certificates.k8s.io/node-csr-MrvfF1bll1SSDFHHOs5bx4Xw3peQ3vz_mNYdeY6hrzA approved
[root@k8s-master01 ~]# kubectl certificate approve node-csr-Tl7t4oF9mp_obWTf5F_55whMrcWStRc3f4TYJtKNTVU
certificatesigningrequest.certificates.k8s.io/node-csr-Tl7t4oF9mp_obWTf5F_55whMrcWStRc3f4TYJtKNTVU approved
[root@k8s-master01 ~]# kubectl certificate approve node-csr-VSttFy1WE5JiBIsfUdKTuCi-EA_DdkLihE1csGCbcXE
certificatesigningrequest.certificates.k8s.io/node-csr-VSttFy1WE5JiBIsfUdKTuCi-EA_DdkLihE1csGCbcXE approved

再次查看 CSR 列表，可以看到节点的加入请求已经被批准

[root@k8s-master01 ~]# kubectl get csr
NAME                                                   AGE     REQUESTOR           CONDITION
node-csr-JxDAOeYmXgaOHWVraQT7ScZ2zt0mj52cC9vA-4WZMUA   35m     kubelet-bootstrap   Approved,Issued
node-csr-MrvfF1bll1SSDFHHOs5bx4Xw3peQ3vz_mNYdeY6hrzA   6m59s   kubelet-bootstrap   Approved,Issued
node-csr-Tl7t4oF9mp_obWTf5F_55whMrcWStRc3f4TYJtKNTVU   6m59s   kubelet-bootstrap   Approved,Issued
node-csr-VSttFy1WE5JiBIsfUdKTuCi-EA_DdkLihE1csGCbcXE   6m59s   kubelet-bootstrap   Approved,Issued

13.从集群剔除node

要删除一个节点前，要先清除掉上面的 pod 然后运行下面的命令删除节点

kubectl drain <node name> --delete-local-data --force --ignore-daemonsets
kubectl delete node <node name>

 如果想要有效删除节点，在节点启动时，重新向集群发送 CSR 请求，还需要在被删除的点节上，删除 CSR 缓存数据

[root@k8s-master02 ~]# ls /home/work/_app/k8s/kubernetes/ssl_cert/
kubelet-client-2019-11-04-20-11-09.pem  kubelet-client-current.pem  kubelet.crt  kubelet.key
[root@k8s-master02 ~]# rm -fr /home/work/_app/k8s/kubernetes/ssl_cert/*

删除完 CSR 缓存数据以后，重启启动 kubelet 就可以在 Master 上收到新的 CSR 请求。

14.给node打标签

查看所有节点状态 ，没打标之前role值默认为none

[root@k8s-master01 ~]# kubectl get node
NAME        STATUS   ROLES    AGE   VERSION
10.0.0.10   Ready    <none>   13h   v1.13.0
10.0.0.11   Ready    <none>   13h   v1.13.0
10.0.0.12   Ready    <none>   13h   v1.13.0
10.0.0.13   Ready    <none>   13h   v1.13.0

k8s-master01 的 Master 打标签

[root@k8s-master01 ~]# kubectl label node 10.0.0.10 node-role.kubernetes.io/master='master'
node/10.0.0.10 labeled

k8s-master02 的 Node打标签

[root@k8s-master01 ~]# kubectl label node 10.0.0.11 node-role.kubernetes.io/master='k8s-master02'
node/10.0.0.11 labeled
[root@k8s-master01 ~]# kubectl label node 10.0.0.11 node-role.kubernetes.io/node='k8s-master02'
node/10.0.0.11 labeled
[root@k8s-master01 ~]# kubectl label node 10.0.0.12 node-role.kubernetes.io/node='k8s-node01'
node/10.0.0.12 labeled
[root@k8s-master01 ~]# kubectl label node 10.0.0.13 node-role.kubernetes.io/node='k8s-node02'
node/10.0.0.13 labeled

[root@k8s-master01 ~]# kubectl get node
NAME        STATUS   ROLES         AGE   VERSION
10.0.0.10   Ready    master        13h   v1.13.0
10.0.0.11   Ready    master,node   13h   v1.13.0
10.0.0.12   Ready    node          13h   v1.13.0
10.0.0.13   Ready    node          13h   v1.13.0

删除掉 k8s-master02 上的 master 标签

[root@k8s-master01 ~]# kubectl label node 10.0.0.11 node-role.kubernetes.io/master-
node/10.0.0.11 labeled
[root@k8s-master01 ~]# kubectl get node
NAME        STATUS   ROLES    AGE   VERSION
10.0.0.10   Ready    master   13h   v1.13.0
10.0.0.11   Ready    node     13h   v1.13.0
10.0.0.12   Ready    node     13h   v1.13.0
10.0.0.13   Ready    node     13h   v1.13.0