【kubernetes系列】kubernetes之使用kubeadm搭建高可用集群

概述

目前来说，kubernetes集群搭建的方式很多，选择一个稳定的适合自己的很重要。目前使用kubeadm方式搭建k8s集群还是很常见的，使用kubeadm搭建可以很简单差不多两条命令就行，也可以稍微复杂一点做一些基础优化，本文将分享一下使用kubeadm搭建集群并做了一定的优化。

安装

本环境将使用centos7.6.1810 的系统安装kubernetes1.19.16版本集群，也可以选择稍高版本的系统和k8s版本。但建议不要使用centos7.4以一下，要不然就自己升级一下内核版本。容器技术依赖于内核技术，低版本系统部署和运行后可能出去一些奇怪的问题。

环境规划

IP	HOSTNAME	role	CPU	Memory
192.168.2.250	vip
192.168.2.140	k8s-m1	master	16	8G
192.168.2.141	k8s-m2	master	16	8G
192.168.2.142	k8s-m3	master	16	8G

说明：

• 所有操作都用root用户进行，系统盘尽量大，要不然就单独修改docker的数据盘，不然到时候镜像多了例如超过85%会被gc回收镜像
• 高可用一般建议大于等于3台的奇数台，本环境我使用的是3台master来做高可用

事前准备(每台机器)

系统层面设置：全新干净系统，只做了网络和dns的相关配置。

• 关闭 所有防火墙和SELinux，否则后续 K8S 挂载目录时可能报错 Permission denied。

systemctl disable --now firewalld NetworkManager
setenforce 0
sed -ri '/^[^#]*SELINUX=/s#=.+$#=disabled#' /etc/selinux/config
ln -sf /usr/share/zoneinfo/Asia/Shanghai /etc/localtime

• 关闭 dnsmasq (可选)
  linux 系统开启了 dnsmasq 后(如 GUI 环境)，将系统 DNS Server 设置为 127.0.0.1，这会导致 docker 容器无法解析域名，需要关闭它

systemctl disable --now dnsmasq

• Kubernetes 建议关闭系统Swap，在所有机器使用以下指令关闭swap并注释掉/etc/fstab中swap的行，不想关闭可以不执行，后面会有应对的配置选项：

swapoff -a && sysctl -w vm.swappiness=0
sed -ri '/^[^#]*swap/s&^&#&' /etc/fstab

• 安装一些基础依赖包和工具

yum install epel-release -y
yum install -y \curl \wget \git \conntrack-tools \psmisc \nfs-utils \jq \socat \bash-completion \ipset \ipvsadm \conntrack \libseccomp \net-tools \crontabs \sysstat \unzip \bind-utils \tcpdump \telnet \lsof \htop

如果集群kube-proxy想使用ipvs模式的话（ipvs效率更高）需要加载以下模块，按照规范使用systemd-modules-load来加载而不是在/etc/rc.local里写modprobe

> /etc/modules-load.d/ipvs.conf
module=(
ip_vs
ip_vs_rr
ip_vs_wrr
ip_vs_sh
nf_conntrack
br_netfilter)
for kernel_module in ${module[@]};do
/sbin/modinfo -F filename $kernel_module |& grep -qv ERROR && echo $kernel_module >> /etc/modules-load.d/ipvs.conf || :
done
systemctl restart systemd-modules-load.service

上面如果systemctl restart 命令报错可以使用systemctl status -l systemd-modules-load.service看看哪个内核模块不能正常加载，然后在/etc/modules-load.d/ipvs.conf里注释掉它再restart试试

• 所有机器需要设定/etc/sysctl.d/k8s.conf的系统参数，目前对ipv6支持不怎么好，所以这里将ipv6也关闭了。

cat <<EOF > /etc/sysctl.d/k8s.conf
net.ipv6.conf.all.disable_ipv6 = 1
net.ipv6.conf.default.disable_ipv6 = 1
net.ipv6.conf.lo.disable_ipv6 = 1
net.ipv4.neigh.default.gc_stale_time = 120
net.ipv4.conf.all.rp_filter = 0
net.ipv4.conf.default.rp_filter = 0
net.ipv4.conf.default.arp_announce = 2
net.ipv4.conf.lo.arp_announce = 2
net.ipv4.conf.all.arp_announce = 2
net.ipv4.ip_forward = 1
net.ipv4.tcp_max_tw_buckets = 5000
net.ipv4.tcp_syncookies = 1
net.ipv4.tcp_max_syn_backlog = 1024
net.ipv4.tcp_synack_retries = 2
# 要求iptables不对bridge的数据进行处理
net.bridge.bridge-nf-call-ip6tables = 1
net.bridge.bridge-nf-call-iptables = 1
net.bridge.bridge-nf-call-arptables = 1
net.netfilter.nf_conntrack_max = 2310720
fs.inotify.max_user_watches=89100
fs.may_detach_mounts = 1
fs.file-max = 52706963
fs.nr_open = 52706963
vm.overcommit_memory=1
vm.panic_on_oom=0
EOF
sysctl --system

• 如果选择关闭swap也要在内核里关闭，不关闭可以不执行

echo 'vm.swappiness = 0' >> /etc/sysctl.d/k8s.conf

• 如果kube-proxy使用ipvs的话为了防止timeout需要设置下tcp参数

cat <<EOF >> /etc/sysctl.d/k8s.conf
# https://github.com/moby/moby/issues/31208 
# ipvsadm -l --timout
# 修复ipvs模式下长连接timeout问题 小于900即可
net.ipv4.tcp_keepalive_time = 600
net.ipv4.tcp_keepalive_intvl = 30
net.ipv4.tcp_keepalive_probes = 10
EOF
sysctl --system

• 修改systemctl启动的最小文件打开数量。关闭ssh方向dns解析

sed -ri 's/^#(DefaultLimitCORE)=/\1=100000/' /etc/systemd/system.conf
sed -ri 's/^#(DefaultLimitNOFILE)=/\1=100000/' /etc/systemd/system.confsed -ri 's/^#(UseDNS )yes/\1no/' /etc/ssh/sshd_config

• 文件最大打开数，按照规范，在子配置文件添加

cat>/etc/security/limits.d/kubernetes.conf<<EOF
*       soft    nproc   131072
*       hard    nproc   131072
*       soft    nofile  131072
*       hard    nofile  131072
root    soft    nproc   131072
root    hard    nproc   131072
root    soft    nofile  131072
root    hard    nofile  131072
EOF

• 集群的HA依赖于时间一致性，安装并配置chrony

yum install -y chrony
cat>/etc/chrony.conf<<EOF
server cn.pool.ntp.org iburst minpoll 4 maxpoll 10
server s1b.time.edu.cn iburst minpoll 4 maxpoll 10
# Ignor source level
stratumweight 0# Record the rate at which the system clock gains/losses time.
driftfile /var/lib/chrony/chrony.drift# This directive enables kernel synchronisation (every 11 minutes) of the
# real-time clock. Note that it can’t be used along with the 'rtcfile' directive.
rtcsync# Allow the system clock to be stepped in the first three updates
# if its offset is larger than 1 second.
makestep 1.0 3# Enable hardware timestamping on all interfaces that support it.
#hwtimestamp *# Increase the minimum number of selectable sources required to adjust
# the system clock.
#minsources 2bindcmdaddress 127.0.0.1#bindcmdaddress ::1# Specify file containing keys for NTP authentication.
keyfile /etc/chrony/chrony.keyslogdir /var/log/chrony# adjust time big than 1 sec will log to file
logchange 1
EOFsystemctl enable --now chronyd

• 修改hostname
  kubelet和kube-proxy上报node信息默认是取hostname的，除非通过--hostname-override指定，这里自行设置hostname，并修改hosts文件。

#按规划进行修改
hostnamectl set-hostname xxx
##所有主机都修改hosts文件
cat >>/etc/hosts << EOF
192.168.2.140 k8s-m1
192.168.2.141 k8s-m2
192.168.2.142 k8s-m3
EOF

• docker官方的内核检查脚本建议(RHEL7/CentOS7: User namespaces disabled; add ‘user_namespace.enable=1’ to boot command line)，使用下面命令开启

grubby --args="user_namespace.enable=1" --update-kernel="$(grubby --default-kernel)"

最后重启系统

reboot

安装docker

选择官方建议的对应版本，可以通过对应版本地址进行查看https://github.com/kubernetes/kubernetes/blob/v1.19.16/build/dependencies.yaml，这里我们选择docker 19.03版本，并使用docker官方的安装脚本进行安装(该脚本支持centos和ubuntu)。

export VERSION=19.03
curl -fsSL "https://get.docker.com/" | bash -s -- --mirror Aliyun

所有机器配置加速源并配置docker的启动参数使用systemd，使用systemd是官方的建议，详见 https://kubernetes.io/docs/setup/cri/

mkdir -p /etc/docker/
cat>/etc/docker/daemon.json<<EOF
{"exec-opts": ["native.cgroupdriver=systemd"],"registry-mirrors": ["https://5sssm2l6.mirror.aliyuncs.com","https://docker.mirrors.ustc.edu.cn/",],"storage-driver": "overlay2","storage-opts": ["overlay2.override_kernel_check=true"],"log-driver": "json-file","log-opts": {"max-size": "100m","max-file": "3"}
}
EOF

Live Restore Enabled这个注意别开，某些极端情况下容器Dead状态之类的必须重启docker daemon才能解决。

• 设置docker开机启动-CentOS安装完成后docker需要手动设置docker命令补全：

yum install -y epel-release bash-completion && \
cp /usr/share/bash-completion/completions/docker /etc/bash_completion.d/

• docker自1.13版起会自动设置iptables的FORWARD默认策略为DROP，这可能会影响Kubernetes集群依赖的报文转发功能，防止FORWARD的DROP策略影响转发,给docker daemon添加下列参数修正，当然暴力点也可以iptables -P FORWARD ACCEPT

mkdir -p /etc/systemd/system/docker.service.d/
cat>/etc/systemd/system/docker.service.d/10-docker.conf<<EOF
[Service]
ExecStartPost=/sbin/iptables -I FORWARD -s 0.0.0.0/0 -j ACCEPT
ExecStopPost=/bin/bash -c '/sbin/iptables -D FORWARD -s 0.0.0.0/0 -j ACCEPT &> /dev/null || :'
EOF

• 启动docker并看下信息是否正常

systemctl enable --now docker
docker info

如果enable docker的时候报错开启debug，如何开见

kubeadm部署

镜像源准备
默认源在国外会无法安装，我们使用国内的镜像源，所有机器都要操作

cat <<EOF >/etc/yum.repos.d/kubernetes.repo
[kubernetes]
name=Kubernetes
baseurl=https://mirrors.aliyun.com/kubernetes/yum/repos/kubernetes-el7-x86_64
enabled=1
gpgcheck=0
EOF

master部分安装相关软件
k8s的node就是kubelet+cri(一般是docker)，kubectl是一个agent读取kubeconfig去访问kube-apiserver来操作集群，kubeadm是部署，所以master节点需要安装三个，node一般不需要kubectl但是yum安装的时候还是会给你安装最新版所以我这里node还是安装了kubectl，不用就行。

yum install -y \kubeadm-1.19.16 \kubectl-1.19.16 \kubelet-1.19.16 \--disableexcludes=kubernetes && \systemctl enable kubelet

node部分安装相关软件

yum install -y \kubeadm-1.19.16 \kubectl-1.19.16 \kubelet-1.19.16 \--disableexcludes=kubernetes && \systemctl enable kubelet

配置kubelet的参数方法(有需要的话)
查看kubelet的systemd文件

$ systemctl cat kubelet
# /usr/lib/systemd/system/kubelet.service
[Unit]
Description=kubelet: The Kubernetes Node Agent
Documentation=https://kubernetes.io/docs/
Wants=network-online.target
After=network-online.target[Service]
ExecStart=/usr/bin/kubelet
Restart=always
StartLimitInterval=0
RestartSec=10[Install]
WantedBy=multi-user.target# /usr/lib/systemd/system/kubelet.service.d/10-kubeadm.conf
# Note: This dropin only works with kubeadm and kubelet v1.11+
[Service]
Environment="KUBELET_KUBECONFIG_ARGS=--bootstrap-kubeconfig=/etc/kubernetes/bootstrap-kubelet.conf --kubeconfig=/etc/kubernetes/kubelet.conf"
Environment="KUBELET_CONFIG_ARGS=--config=/var/lib/kubelet/config.yaml"
# This is a file that "kubeadm init" and "kubeadm join" generates at runtime, populating the KUBELET_KUBEADM_ARGS variable dynamically
EnvironmentFile=-/var/lib/kubelet/kubeadm-flags.env
# This is a file that the user can use for overrides of the kubelet args as a last resort. Preferably, the user should use
# the .NodeRegistration.KubeletExtraArgs object in the configuration files instead. KUBELET_EXTRA_ARGS should be sourced from this file.
EnvironmentFile=-/etc/sysconfig/kubelet
Environment="KUBELET_CGROUP_ARGS=--cgroup-driver=systemd"
ExecStart=
ExecStart=/usr/bin/kubelet $KUBELET_KUBECONFIG_ARGS $KUBELET_CONFIG_ARGS $KUBELET_KUBEADM_ARGS $KUBELET_EXTRA_ARGS

我们可以看到/etc/sysconfig/kubelet和/var/lib/kubelet/kubeadm-flags.env都是EnvironmentFile，通过可以通过注释查看每个文件中如何设置运行参数。如/var/lib/kubelet/kubeadm-flags.env 中应该通过KUBELET_EXTRA_ARGS来给kubelet配置运行参数，如下。具体参数可以通过kubelet --help查看，参数很多，很多不了解的最好保持默认。

cat >/etc/sysconfig/kubelet<<EOF
KUBELET_EXTRA_ARGS="--aa=bb --xx=yy"
EOF
第二个EnvironmentFile 文件/var/lib/kubelet/kubeadm-flags.env也一样

haproxy和keepalived的安装

三台master都需要安装
haproxy配置文件
注意，三个节点上haproxy.conf的配置文件内容其实是一样的

cat <<EOF > /etc/haproxy/haproxy.cfg
globalmaxconn  2000ulimit-n  16384log  127.0.0.1 local0 errstats timeout 30sdefaultslog globalmode  httpoption  httplogtimeout connect 5000timeout client  50000timeout server  50000timeout http-request 15stimeout http-keep-alive 15sfrontend monitor-inbind *:33305mode httpoption httplogmonitor-uri /monitorlisten statsbind    *:8006mode    httpstats   enablestats   hide-versionstats   uri       /statsstats   refresh   30sstats   realm     Haproxy\ Statisticsstats   auth      admin:adminfrontend k8s-apibind 0.0.0.0:8443bind 127.0.0.1:8443mode tcpoption tcplogtcp-request inspect-delay 5sdefault_backend k8s-apibackend k8s-apimode tcpoption tcplogoption httpchk GET /healthzhttp-check expect string okbalance roundrobindefault-server inter 10s downinter 5s rise 2 fall 2 slowstart 60s maxconn 250 maxqueue 256 weight 100server  api1  192.168.2.140:6443  check check-ssl verify noneserver  api2  192.168.2.141:6443  check check-ssl verify noneserver  api3  192.168.2.142:6443  check check-ssl verify none
EOF

keepalived配置文件
注意，keepalived.conf的配置文件内容是有区别的。

cat <<EOF > /etc/keepalived/keepalived.conf
global_defs {enable_script_security
}vrrp_script haproxy-check {user rootscript "/bin/bash /etc/keepalived/check_haproxy.sh"interval 3weight -2fall 10rise 2
}vrrp_instance haproxy-vip {state MASTER  ##注意修改，其中主为MASTER，从为BACKUPpriority 100 ##注意修改，数字越大，优先级越高，主>从interface eth0   #注意实际环境中网卡名字，有的是ens*啥的virtual_router_id 48 #所有节点的id要一致advert_int 3unicast_src_ip 192.168.2.140           # 本机IPunicast_peer {192.168.2.141      # 对端IP192.168.2.142      # 对端IP}virtual_ipaddress {192.168.2.250/24   # VIP地址}track_script {haproxy-check}
}
EOF

说明：keeaplived这里需要注意，默认keepalived是采用的组播方式，加上unicast_peer参数后是单播方式，三台keepalived配置文件不一样unicast_src_ipc参数写当前节点IP，unicast_peer参数写另外两个节点IP地址。其他按说明修改

keepalived 健康检查脚本

 cat <<'EOF'> /etc/keepalived/check_haproxy.sh
#!/bin/bash
VIRTUAL_IP=192.168.2.250.250errorExit() {echo "*** $*" 1>&2exit 1
}if ip addr | grep -q $VIRTUAL_IP ; thencurl -s --max-time 2 --insecure https://${VIRTUAL_IP}:8443/healthz  -o /dev/null || errorExit "Error GET https://${VIRTUAL_IP}:8443/healthz"
fi
EOF

部署外部Etcd服务

openssl 证书配置文件，注意IP地址书写正确

mkdir -p /etc/kubernetes/pki/etcd
cd /etc/kubernetes/pkicat <<EOF> /etc/kubernetes/pki/openssl.cnf
[ req ]
default_bits = 2048
default_md = sha256
distinguished_name = req_distinguished_name[req_distinguished_name][ v3_ca ]
basicConstraints = critical, CA:TRUE
keyUsage = critical, digitalSignature, keyEncipherment, keyCertSign[ v3_req_server ]
basicConstraints = CA:FALSE
keyUsage = critical, digitalSignature, keyEncipherment
extendedKeyUsage = serverAuth[ v3_req_client ]
basicConstraints = CA:FALSE
keyUsage = critical, digitalSignature, keyEncipherment
extendedKeyUsage = clientAuth[ v3_req_apiserver ]
basicConstraints = CA:FALSE
keyUsage = critical, digitalSignature, keyEncipherment
extendedKeyUsage = serverAuth
subjectAltName = @alt_names_cluster[ v3_req_etcd ]
basicConstraints = CA:FALSE
keyUsage = critical, digitalSignature, keyEncipherment
extendedKeyUsage = serverAuth, clientAuth
subjectAltName = @alt_names_etcd[ alt_names_cluster ]
DNS.1 = kubernetes
DNS.2 = kubernetes.default
DNS.3 = kubernetes.default.svc
DNS.4 = kubernetes.default.svc.cluster.local
DNS.5 = k8s-m1
DNS.6 = k8s-m2
DNS.7 = k8s-m3
DNS.8 = localhost
IP.1 = 10.96.0.1
IP.2 = 127.0.0.1
IP.3 = 192.168.2.140  
IP.4 = 192.168.2.141
IP.5 = 192.168.2.142
IP.6 = 192.168.2.250[ alt_names_etcd ]
DNS.1 = localhost
DNS.2 = k8s-m1
DNS.3 = k8s-m2
DNS.4 = k8s-m3
IP.1 = 192.168.2.140
IP.2 = 192.168.2.141
IP.3 = 192.168.2.142
IP.4 = 127.0.0.1
EOF

根据上面的配置文件，生成各服务所需证书，并下发到其他节点。其他节点就不需要在进行证书生成
#生成证书，有效期10000d

openssl genrsa -out etcd/ca.key 2048
openssl req -x509 -new -nodes -key etcd/ca.key -config openssl.cnf -subj "/CN=etcd-ca" -extensions v3_ca -out etcd/ca.crt -days 10000
openssl genrsa -out apiserver-etcd-client.key 2048
openssl req -new -key apiserver-etcd-client.key -subj "/CN=apiserver-etcd-client/O=system:masters" -out apiserver-etcd-client.csr
openssl x509 -in apiserver-etcd-client.csr -req -CA etcd/ca.crt -CAkey etcd/ca.key -CAcreateserial -extensions v3_req_etcd -extfile openssl.cnf -out apiserver-etcd-client.crt -days 10000
openssl genrsa -out etcd/server.key 2048
openssl req -new -key etcd/server.key -subj "/CN=etcd-server" -out etcd/server.csr
openssl x509 -in etcd/server.csr -req -CA etcd/ca.crt -CAkey etcd/ca.key -CAcreateserial -extensions v3_req_etcd -extfile openssl.cnf -out etcd/server.crt -days 10000
openssl genrsa -out etcd/peer.key 2048
openssl req -new -key etcd/peer.key -subj "/CN=etcd-peer" -out etcd/peer.csr
openssl x509 -in etcd/peer.csr -req -CA etcd/ca.crt -CAkey etcd/ca.key -CAcreateserial -extensions v3_req_etcd -extfile openssl.cnf -out etcd/peer.crt -days 10000
openssl genrsa -out etcd/healthcheck-client.key 2048
openssl req -new -key etcd/healthcheck-client.key -subj "/CN=etcd-client" -out etcd/healthcheck-client.csr
openssl x509 -in etcd/healthcheck-client.csr -req -CA etcd/ca.crt -CAkey etcd/ca.key -CAcreateserial -extensions v3_req_etcd -extfile openssl.cnf -out etcd/healthcheck-client.crt -days 10000scp -r /etc/kubernetes root@k8s-m2:/etc
scp -r /etc/kubernetes root@k8s-m3:/etc

下载部署，其他节点类似：
所需etcd对应版本可以通过https://github.com/kubernetes/kubernetes/blob/v1.19.16/build/dependencies.yaml查看

mkdir -p /var/lib/etcd
ETCD_VER=v3.4.13
wget https://storage.googleapis.com/etcd/${ETCD_VER}/etcd-${ETCD_VER}-linux-amd64.tar.gz -O etcd-${ETCD_VER}-linux-amd64.tar.gz
tar xf etcd-${ETCD_VER}-linux-amd64.tar.gz  --strip-components=1 -C /usr/local/bin etcd-${ETCD_VER}-linux-amd64/{etcd,etcdctl}

设置 unit file 并启动 etcd,其他节点修改对应 ETCD_NAME 为 etcd1 和 etcd2，ip 改为节点 IP。

ETCD_NAME=etcd0
ETCD_IP="192.168.2.140"
ETCD_IPS=(192.168.2.140 192.168.2.141 192.168.2.142)cat<<EOF> /usr/lib/systemd/system/etcd.service 
[Unit]
Description=etcd
Documentation=https://coreos.com/etcd/docs/latest/
After=network.target[Service]
Type=notify
WorkingDirectory=/var/lib/etcd
ExecStart=/usr/bin/etcd \\--name=${ETCD_NAME} \\--data-dir=/var/lib/etcd \\--listen-client-urls=https://127.0.0.1:2379,https://${ETCD_IP}:2379 \\--advertise-client-urls=https://${ETCD_IP}:2379 \\--listen-peer-urls=https://${ETCD_IP}:2380 \\--initial-advertise-peer-urls=https://${ETCD_IP}:2380 \\--cert-file=/etc/kubernetes/pki/etcd/server.crt \\--key-file=/etc/kubernetes/pki/etcd/server.key \\--client-cert-auth \\--trusted-ca-file=/etc/kubernetes/pki/etcd/ca.crt \\--peer-cert-file=/etc/kubernetes/pki/etcd/peer.crt \\--peer-key-file=/etc/kubernetes/pki/etcd/peer.key \\--peer-client-cert-auth \\--peer-trusted-ca-file=/etc/kubernetes/pki/etcd/ca.crt \\--initial-cluster=etcd0=https://${ETCD_IPS[0]}:2380,etcd1=https://${ETCD_IPS[1]}:2380,etcd2=https://${ETCD_IPS[2]}:2380 \\--initial-cluster-token=my-etcd-token \\--initial-cluster-state=new \\--heartbeat-interval 1000 \\--election-timeout 5000Restart=always
RestartSec=10s
LimitNOFILE=65535[Install]
WantedBy=multi-user.target
EOFsystemctl daemon-reload
systemctl restart etcd
systemctl enable etcdcat <<EOF > /etc/profile.d/etcd.sh
alias etcd_v2='etcdctl --cert-file /etc/kubernetes/pki/etcd/healthcheck-client.crt \--key-file /etc/kubernetes/pki/etcd/healthcheck-client.key  \--ca-file /etc/kubernetes/pki/etcd/ca.crt  \--endpoints https://192.168.2.140:2379,https://192.168.2.141:2379,https://192.168.2.142:2379'alias etcd_v3='ETCDCTL_API=3 \etcdctl   \--cert /etc/kubernetes/pki/etcd/healthcheck-client.crt \--key /etc/kubernetes/pki/etcd/healthcheck-client.key \--cacert /etc/kubernetes/pki/etcd/ca.crt \--endpoints https://192.168.2.140:2379,https://192.168.2.141:2379,https://192.168.2.142:2379'
EOF
source /etc/profile.d/etcd.sh
etcd_v3 version
etcdctl version: 3.4.13
API version: 3.4
etcd_v3 --write-out=table endpoint status[root@k8s-m1 ~]# etcd_v3 --write-out=table endpoint status
+----------------------------+------------------+---------+---------+-----------+-----------+------------+
|          ENDPOINT          |        ID        | VERSION | DB SIZE | IS LEADER | RAFT TERM | RAFT INDEX |
+----------------------------+------------------+---------+---------+-----------+-----------+------------+
| https://192.168.2.140:2379 | 19df3a9852e0345a |  3.4.13 |   24 MB |     false |    267804 |  108091120 |
| https://192.168.2.141:2379 | 66d402f1ef2c996e |  3.4.13 |   24 MB |      true |    267804 |  108091120 |
| https://192.168.2.142:2379 | 3bb3629d60bef3f6 |  3.4.13 |   24 MB |     false |    267804 |  108091121 |
+----------------------------+------------------+---------+---------+-----------+-----------+------------+```

k8s集群安装(第一个master上配置)

打印默认init的配置信息

kubeadm config print init-defaults > initconfig.yaml

我们看下默认init的集群参数

apiVersion: kubeadm.k8s.io/v1beta2
bootstrapTokens:
- groups:- system:bootstrappers:kubeadm:default-node-tokentoken: abcdef.0123456789abcdefttl: 24h0m0susages:- signing- authentication
kind: InitConfiguration
localAPIEndpoint:advertiseAddress: 1.2.3.4bindPort: 6443
nodeRegistration:criSocket: /var/run/dockershim.sockname: k8s-m1taints:- effect: NoSchedulekey: node-role.kubernetes.io/master
---
apiServer:timeoutForControlPlane: 4m0s
apiVersion: kubeadm.k8s.io/v1beta2
certificatesDir: /etc/kubernetes/pki
clusterName: kubernetes
controllerManager: {}
dns:type: CoreDNS
etcd:local:dataDir: /var/lib/etcd
imageRepository: k8s.gcr.io
kind: ClusterConfiguration
kubernetesVersion: v1.19.0
networking:dnsDomain: cluster.localserviceSubnet: 10.96.0.0/12
scheduler: {}

我们主要关注和保留ClusterConfiguration的段，然后修改下，可以参考下列的v1beta2文档，如果是低版本可能是v1beta1，某些字段和新的是不一样的，自行查找godoc看
https://godoc.org/k8s.io/kubernetes/cmd/kubeadm/app/apis/kubeadm/v1beta2#hdr-Basics
https://godoc.org/k8s.io/kubernetes/cmd/kubeadm/app/apis/kubeadm/v1beta2
https://godoc.org/k8s.io/kubernetes/cmd/kubeadm/app/apis/kubeadm/v1beta2#pkg-constants
https://godoc.org/k8s.io/kubernetes/cmd/kubeadm/app/apis/kubeadm/v1beta2#ClusterConfiguration

controlPlaneEndpoint是规划的vip地址，下面是最终的yaml

apiVersion: kubeadm.k8s.io/v1beta2
kind: ClusterConfiguration
imageRepository: registry.aliyuncs.com/google_containers
kubernetesVersion: v1.19.16
certificatesDir: /etc/kubernetes/pki
clusterName: kubernetes
networking: dnsDomain: cluster.localserviceSubnet: 10.96.0.0/12podSubnet: 10.244.0.0/16
controlPlaneEndpoint: 192.168.2.250:8443 # 单个master的话写master的ip或者不写，端口是haproxy运行的端口
apiServer: # https://godoc.org/k8s.io/kubernetes/cmd/kubeadm/app/apis/kubeadm/v1beta3#APIServertimeoutForControlPlane: 4m0sextraArgs:authorization-mode: "Node,RBAC"enable-admission-plugins: "NamespaceLifecycle,LimitRanger,ServiceAccount,PersistentVolumeClaimResize,DefaultStorageClass,DefaultTolerationSeconds,NodeRestriction,MutatingAdmissionWebhook,ValidatingAdmissionWebhook,ResourceQuota,Priority"runtime-config: api/all=truestorage-backend: etcd3certSANs:- 127.0.0.1 # 多个master的时候负载均衡出问题了能够快速使用localhost调试- localhost- 192.168.2.140- 192.168.2.141- 192.168.2.142- k8s-m1- k8s-m2- k8s-m3extraVolumes:- hostPath: /etc/localtimemountPath: /etc/localtimename: localtimereadOnly: true
controllerManager: extraArgs:bind-address: "0.0.0.0"extraVolumes:- hostPath: /etc/localtimemountPath: /etc/localtimename: localtimereadOnly: true
scheduler: extraArgs:bind-address: "0.0.0.0"extraVolumes:- hostPath: /etc/localtimemountPath: /etc/localtimename: localtimereadOnly: true
dns: {}
etcd: external:endpoints:- https://192.168.2.140:2379- https://192.168.2.141:2379- https://192.168.2.142:2379caFile: /etc/kubernetes/pki/etcd/ca.crtcertFile: /etc/kubernetes/pki/apiserver-etcd-client.crtkeyFile: /etc/kubernetes/pki/apiserver-etcd-client.key
---
apiVersion: kubeproxy.config.k8s.io/v1alpha1
kind: KubeProxyConfiguration 
mode: ipvs # or iptables
ipvs:excludeCIDRs: nullminSyncPeriod: 0sscheduler: "rr" # 调度算法strictARP: falsesyncPeriod: 15s
iptables:masqueradeAll: truemasqueradeBit: 14minSyncPeriod: 0ssyncPeriod: 30s
---
apiVersion: kubelet.config.k8s.io/v1beta1
kind: KubeletConfiguration 
cgroupDriver: systemd
failSwapOn: true # 如果开启swap则设置为false

• swap的话看最后一行，单台master的话把controlPlaneEndpoint的值改为第一个master的ip
• kubectl get cs 查看组件状态发现controllerManager和scheduler 状态Unhealthy 删除/etc/kubernetes/manifests/目录下对应文件中–port=0的配置即可，每个master节点都需要删除
• 检查文件是否错误，忽略warning，错误的话会抛出error，没有错误则会输出到包含字符串kubeadm join 等添加节点的信息

kubeadm init --config initconfig.yaml --dry-run

#检查镜像是否正确
kubeadm config images list --config initconfig.yaml
#预先拉取镜像
kubeadm config images pull --config initconfig.yaml # 下面是输出
[config/images] Pulled registry.aliyuncs.com/google_containers/kube-apiserver:v1.19.16
[config/images] Pulled registry.aliyuncs.com/google_containers/kube-controller-manager:v1.19.16
[config/images] Pulled registry.aliyuncs.com/google_containers/kube-scheduler:v1.19.16
[config/images] Pulled registry.aliyuncs.com/google_containers/kube-proxy:v1.19.16
[config/images] Pulled registry.aliyuncs.com/google_containers/pause:3.2
[config/images] Pulled registry.aliyuncs.com/google_containers/coredns:1.7.0

kubeadm 初始化

下面init只在第一个master上面操作

kubeadm init --config initconfig.yaml 

初始化的时候可以添加参数–upload-certs， 作用为将相关的证书直接上传到etcd中保存，这样省去我们手动分发证书的过程。
初始化完成后记住init后打印的token，复制kubectl的kubeconfig，kubectl的kubeconfig路径默认是~/.kube/config

mkdir -p $HOME/.kube
sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
sudo chown $(id -u):$(id -g) $HOME/.kube/config

init的yaml信息实际上会存在集群的configmap里，我们可以随时查看，该yaml在其他node和master join的时候会使用到

kubectl -n kube-system get cm kubeadm-config -o yaml

配置其他master的k8s管理组件(某些低版本不支持上传证书的时候操作需手动拷贝证书)

第一个master上拷贝ca证书到其他master节点上

scp -r /etc/kubernetes/pki root@k8s-m2:/etc/kubernetes/
scp -r /etc/kubernetes/pki root@k8s-m3:/etc/kubernetes/

其他master join进来

kubeadm join 192.168.2.250.250:8443 \--token xxx.zzzzzzzzz     \--discovery-token-ca-cert-hash sha256:xxxxxxxxxxx --control-plane

通过下列命令可以获取sha256的值

openssl x509 -pubkey -in \/etc/kubernetes/pki/ca.crt | \openssl rsa -pubin -outform der 2>/dev/null | \openssl dgst -sha256 -hex | sed 's/^.* //'

• 如果集群在init时使用了 --upload-certs 参数将相关的证书直接上传到etcd中保存，则其他master在加入时需要使用 --certificate-key 参数（某些低版本可能不支持）。
• token忘记的话可以kubeadm token list查看，没有的话可以通过kubeadm token create创建。在高版本可以直接使用kubeadm token create --print-join-command来创建添加节点的命令，某些老版本可能不确定支持–print-join-command这个选项，不支持的话就不带–print-join-command选项创建token。
• 通过将参数 --upload-certs 添加到 kubeadm init，你可以将控制平面证书临时上传到集群中的 Secret 请注意此 Secret 将在 2小时后自动过期。证书使用 32 字节密钥加密，可以使用 --certificate-key 指定。
  以下阶段命令可用于证书到期后重新上传证书：

kubeadm init phase upload-certs --upload-certs --certificate-key=SOME_VALUE

如果未将参数 --certificate-key 传递给 kubeadm init 和 kubeadm init phase upload-certs， 则会自动生成一个新密钥。

以下命令可用于按需生成新密钥：

kubeadm alpha certs certificate-key

设置kubectl的补全脚本

kubectl completion bash > /etc/bash_completion.d/kubectl

获取节点状态信息

[root@k8s-m1 ~]# kubectl get nodes 
NAME     STATUS   ROLES    AGE    VERSION
k8s-m1   Ready    master   1d   v1.19.16
k8s-m2   Ready    master   1d   v1.19.16
k8s-m3   Ready    master   1d   v1.19.16[root@k8s-m1 ~]# kubectl get cs
Warning: v1 ComponentStatus is deprecated in v1.19+
NAME                 STATUS    MESSAGE             ERROR
controller-manager   Healthy   ok                  
scheduler            Healthy   ok                  
etcd-1               Healthy   {"health":"true"}   
etcd-2               Healthy   {"health":"true"}   
etcd-0               Healthy   {"health":"true"} 

addon(此章开始到结尾选取任意一个master上执行)

到此，集群还不能真正使用，因为集群的网络组件是以插件的方式部署，在这里我选用了常用的flannel，后面将分享使用其他网络组件，如calico。

#直接apply就行，或者先用wget先下载下来也可以，多尝试几次就能下载下来
kubectl apply -f https://raw.githubusercontent.com/coreos/flannel/master/Documentation/kube-flannel.yml

验证集群可用性
等kube-system空间下的pod都处于running状态后再测试集群的可用性

[root@k8s-m1 k8s-total]# cat test-k8s.yaml 
apiVersion: apps/v1
kind: Deployment
metadata:name: my-nginx
spec:selector:matchLabels:tier: frontendreplicas: 1template:metadata:labels:tier: frontendspec:containers:- name: nginximage: nginxports:- containerPort: 80
---
apiVersion: v1
kind: Service
metadata:name: nginx
spec:selector:tier: frontendports:- protocol: TCPport: 80targetPort: 80
---
apiVersion: v1
kind: Pod
metadata:name: busybox
spec:containers:- name: busyboximage: busybox:1.28.4command:- sleep- "3600"imagePullPolicy: IfNotPresentrestartPolicy: Always
##注意busybox的版本要1.28.4之前的版本，不然解析有问题

[root@k8s-m1 k8s-total]# kubectl get po,svc
NAME                            READY   STATUS    RESTARTS   AGE     IP           NODE     NOMINATED NODE   READINESS GATES
pod/busybox                     1/1     Running   0          5m23s   10.244.0.4   k8s-m2   <none>           <none>
pod/my-nginx-5b8555d6b8-vxcsj   1/1     Running   0          20m     10.244.2.3   k8s-m1   <none>           <none>NAME                           TYPE        CLUSTER-IP       EXTERNAL-IP   PORT(S)    AGE     SELECTOR
service/kubernetes             ClusterIP   10.96.0.1        <none>        443/TCP    436d    <none>service/nginx                  ClusterIP   10.107.34.204    <none>        80/TCP     16m
service/nginx                  ClusterIP   10.107.34.204    <none>        80/TCP     20m     tier=frontend

验证集群dns，使用nslookup或者dig都可以

[root@k8s-m1 k8s-total]#  kubectl exec -ti busybox -- nslookup kubernetes
Server:    10.96.0.10
Address 1: 10.96.0.10 kube-dns.kube-system.svc.cluster.localName:      kubernetes
Address 1: 10.96.0.1 kubernetes.default.svc.cluster.local[root@k8s-m1 k8s-total]# dig -t A kubernetes.default.svc.cluster.local.  @10.96.0.10
#日常检查其他服务能否正常解析时也可以使用此命令

在master上curl nginx的svc的ip出现nginx的index内容即集群正常，例如我的nginx svc ip是10.107.34.204

[root@k8s-m1 k8s-total]# curl 10.107.34.204  -s
<!DOCTYPE html>
<html>
<head>
<title>Welcome to nginx!</title>
<style>
html { color-scheme: light dark; }
body { width: 35em; margin: 0 auto;
font-family: Tahoma, Verdana, Arial, sans-serif; }
</style>
</head>
<body>
<h1>Welcome to nginx!</h1>
<p>If you see this page, the nginx web server is successfully installed and
working. Further configuration is required.</p><p>For online documentation and support please refer to
<a href="http://nginx.org/">nginx.org</a>.<br/>
Commercial support is available at
<a href="http://nginx.com/">nginx.com</a>.</p><p><em>Thank you for using nginx.</em></p>
</body>
</html>

日志管理

以kube-apiserver为例将日志挂载出来方便管理（kube-controller-manager、kube-scheduler组件服务类似）
修改以下三处

#启动参数
spec:containers:- command:- --logtostderr=false- --log-dir=/var/log/kubernetes/kube-apiserver- --v=2#mount地址
volumeMounts:
- mountPath: /var/log/kubernetes/kube-apiservername: k8s-logs#被挂载券设置
volumes:
- hostPath:path: /var/log/kubernetes/kube-apiservertype: DirectoryOrCreatename: k8s-logs

kubelet日志（kubelet服务是用systemctl进行管理的非容器管理所以不用挂载），可以通过设置文件存放目录直接进行修改。

[root@k8s-m1 manifests]# vim /etc/sysconfig/kubelet
--v=2 --logtostderr=false --log-dir=/var/log/kubernetes/kubelet

更多关于kubernetes的知识分享，请前往博客主页。编写过程中，难免出现差错，敬请指出