IT苦工指南 | Kubernetes v1.8.x全手动安装

IT苦工指南 | Kubernetes v1.8.x全手动安装

最近,有部分用户飘了……

觉得Rainbond提供的既简洁、又易用、而且生产就绪的Kubernets体验不过瘾……

想要挑战一下Kubernetes全手动部署……

并在凌晨一点拨通了客服小哥的电话……

因此本着不重复造轮子并且关爱客服小哥身心健康的主张,我们搬来了Kairen的精彩教程——

开始

Kubernetes官方提供了多种安装方式Picking the right solution,本文将以全手动安装方式来部署Kubernetes v1.8.x版本,学习和了解Kubernetes的构建流程。

版本明细:

  • Kubernetes v1.8.6
  • CNI v0.6.0
  • Etcd v3.2.9
  • Calico v2.6.2
  • Docker v17.10.0-ce

准备

系统:ubuntu 16.xcentos 7.x 节点:

  • 172.16.35.12 / master1 / 1 CPU / 2G
  • 172.16.35.10 / node1 / 1 CPU / 2G
  • 172.16.35.11 / node2 / 1 CPU / 2G

master为主要控制节点和部署节点,node为应用运行节点 所有操作均为root

安装前需确认以下事项:

  • 确认所有节点之间网络互通,master1 SSH登入其他节点为passwordless
  • 确认防火墙和SELinux已关闭,如centos:
$ systemctl stop firewalld && systemctl disable firewalld
$ setenforce 0
$ vim /etc/selinux/config
SELINUX=disabled  
  • 所有节点需要设置/etc/host解析到所有主机
...
172.16.35.10 node1  
172.16.35.11 node2  
172.16.35.12 master1  
  • 所有节点都需要安装docker
$ curl -fsSL "https://get.docker.com/" | sh 

注意:centos安装docker完成后需要执行:

$ systemctl enable docker && systemctl start docker 

编辑/lib/systemd/system/docker.service,在ExecStart=..加入:

ExecStartPost=/sbin/iptables -A FORWARD -s 0.0.0.0/0 -j ACCEPT  

完成后重启docker服务:

$ systemctl daemon-reload && systemctl restart docker 
  • 所有节点都需要设定/etc/sysctl.d/k8s.conf系统参数
$ cat <<EOF > /etc/sysctl.d/k8s.conf
net.ipv4.ip_forward = 1  
net.bridge.bridge-nf-call-ip6tables = 1  
net.bridge.bridge-nf-call-iptables = 1  
EOF

$ sysctl -p /etc/sysctl.d/k8s.conf
  • 在master1安装CFSSL工具,用来建立TLS certificates
$ export CFSSL_URL="https://pkg.cfssl.org/R1.2"
$ wget "${CFSSL_URL}/cfssl_linux-amd64" -O /usr/local/bin/cfssl
$ wget "${CFSSL_URL}/cfssljson_linux-amd64" -O /usr/local/bin/cfssljson
$ chmod +x /usr/local/bin/cfssl /usr/local/bin/cfssljson

Etcd

安装Kubernetes之前,我们需要完成一些必要的系统配置,高可用共享配置和服务发现存储Etcd便是其中的重要一环,节点会从Etcd中获取所需数据。

建立集群CA和certificates

这里需要生成client和server各组件certificate,代替kubernetes admin user生成client证书。

首先在master1建立/etc/etcd/ssl目录,而后进入目录进行以下操作:

$ mkdir -p /etc/etcd/ssl && cd /etc/etcd/ssl
$ export PKI_URL="https://kairen.github.io/files/manual-v1.8/pki"

下载ca-config.jsonetcd-ca-csr.json

$ wget "${PKI_URL}/ca-config.json" "${PKI_URL}/etcd-ca-csr.json"
$ cfssl gencert -initca etcd-ca-csr.json | cfssljson -bare etcd-ca
$ ls etcd-ca*.pem
etcd-ca-key.pem  etcd-ca.pem  

下载etcd-csr.json并生成Etcd certificate证书:

$ wget "${PKI_URL}/etcd-csr.json"
$ cfssl gencert \
  -ca=etcd-ca.pem \
  -ca-key=etcd-ca-key.pem \
  -config=ca-config.json \
  -profile=kubernetes \
  etcd-csr.json | cfssljson -bare etcd

$ ls etcd*.pem
etcd-ca-key.pem  etcd-ca.pem  etcd-key.pem  etcd.pem  

如果节点IP不同,需要修改etcd-csr.json的hosts

完成后删除不必要的文件:

$ rm -rf *.json 

并确认/etc/etcd/ssl包含:

$ ls /etc/etcd/ssl
etcd-ca.csr  etcd-ca-key.pem  etcd-ca.pem  etcd.csr  etcd-key.pem  etcd.pem  

Etcd的安装和设置

首先在master1节点下载Etcd,解压到/opt安装:

$ export ETCD_URL="https://github.com/coreos/etcd/releases/download"
$ cd && wget -qO- --show-progress "${ETCD_URL}/v3.2.9/etcd-v3.2.9-linux-amd64.tar.gz" | tar -zx
$ mv etcd-v3.2.9-linux-amd64/etcd* /usr/local/bin/ && rm -rf etcd-v3.2.9-linux-amd64

完成后新建Etcd Group和User,并设定Etcd目录:

$ groupadd etcd && useradd -c "Etcd user" -g etcd -s /sbin/nologin -r etcd

下载etcd相关配置,我们将来管理Etcd:

$ export ETCD_CONF_URL="https://kairen.github.io/files/manual-v1.8/master"
$ wget "${ETCD_CONF_URL}/etcd.conf" -O /etc/etcd/etcd.conf
$ wget "${ETCD_CONF_URL}/etcd.service" -O /lib/systemd/system/etcd.service

如果没用本文准备部分的IP,请用自己的IP代替172.16.35.12

建立var 存放数据,然后启动Etcd服务:

$ mkdir -p /var/lib/etcd && chown etcd:etcd -R /var/lib/etcd /etc/etcd
$ systemctl enable etcd.service && systemctl start etcd.service 

通过以下命令验证:

$ export CA="/etc/etcd/ssl"
$ ETCDCTL_API=3 etcdctl \
    --cacert=${CA}/etcd-ca.pem \
    --cert=${CA}/etcd.pem \
    --key=${CA}/etcd-key.pem \
    --endpoints="https://172.16.35.12:2379" \
    endpoint health
# output
https://172.16.35.12:2379 is healthy: successfully committed proposal: took = 641.36µs  

Kubernetes Master

Master是Kubernetes的大总管,通过apiserverController manager以及Scheduler管理所有节点。

本部分将下载Kubernetes并安装到master1节点上,然后生成相关TLS certificates和CA,供集群组件使用。

下载Kubernetes组件

# Download Kubernetes
$ export KUBE_URL="https://storage.googleapis.com/kubernetes-release/release/v1.8.6/bin/linux/amd64"
$ wget "${KUBE_URL}/kubelet" -O /usr/local/bin/kubelet
$ wget "${KUBE_URL}/kubectl" -O /usr/local/bin/kubectl
$ chmod +x /usr/local/bin/kubelet /usr/local/bin/kubectl

# Download CNI
$ mkdir -p /opt/cni/bin && cd /opt/cni/bin
$ export CNI_URL="https://github.com/containernetworking/plugins/releases/download"
$ wget -qO- --show-progress "${CNI_URL}/v0.6.0/cni-plugins-amd64-v0.6.0.tgz" | tar -zx

建立集群CA和certificates

与Etcd部分原理一样,操作也大相径庭,首先在master1建立pki目录,并进入目录执行:

$ mkdir -p /etc/kubernetes/pki && cd /etc/kubernetes/pki
$ export PKI_URL="https://kairen.github.io/files/manual-v1.8/pki"
$ export KUBE_APISERVER="https://172.16.35.12:6443"

下载ca-config.jsonetcd-ca-csr.json

$ wget "${PKI_URL}/ca-config.json" "${PKI_URL}/ca-csr.json"
$ cfssl gencert -initca ca-csr.json | cfssljson -bare ca
$ ls ca*.pem
ca-key.pem  ca.pem  
API server certificate

下载apiserver-csr.json,生成kube-apiserver certificate证书:

$ wget "${PKI_URL}/apiserver-csr.json"
$ cfssl gencert \
  -ca=ca.pem \
  -ca-key=ca-key.pem \
  -config=ca-config.json \
  -hostname=10.96.0.1,172.16.35.12,127.0.0.1,kubernetes.default \
  -profile=kubernetes \
  apiserver-csr.json | cfssljson -bare apiserver

$ ls apiserver*.pem
apiserver-key.pem  apiserver.pem  

如果节点IP不同,需要修改-hostname

Front proxy certificate

下载front-proxy-ca-csr.json,生成Front proxy CA,Front proxy主要用在API aggregator上:

$ wget "${PKI_URL}/front-proxy-ca-csr.json"
$ cfssl gencert \
  -initca front-proxy-ca-csr.json | cfssljson -bare front-proxy-ca

$ ls front-proxy-ca*.pem
front-proxy-ca-key.pem  front-proxy-ca.pem  

下载front-proxy-client-csr.json,生成front-proxy-client证书:

$ wget "${PKI_URL}/front-proxy-client-csr.json"
$ cfssl gencert \
  -ca=front-proxy-ca.pem \
  -ca-key=front-proxy-ca-key.pem \
  -config=ca-config.json \
  -profile=kubernetes \
  front-proxy-client-csr.json | cfssljson -bare front-proxy-client

$ ls front-proxy-client*.pem
front-proxy-client-key.pem  front-proxy-client.pem  
Bootstrap Token

手工方式生成CA非常麻烦,只适合少量机器,每次签证时都需要绑定Node IP,随着机器增加会带来很多的不便,因此这里使用TLS Bootstrapping的方式来进行授权,由apiserver自动为符合条件的Node发送证书授权加入集群。

做法是在kubelet启动时,向kuber-apiserver传送TLS Bootstrapping请求,而kube-apiserver验证kubelet请求的token是否与设定的一样,如果一样则自动生成Kuberlet证书和密钥。具体作法可以参考TLS bootstrapping

首先生成BOOTSTRAP_TOKEN,并建立bootstrap.conf的kubeconfig:

$ export BOOTSTRAP_TOKEN=$(head -c 16 /dev/urandom | od -An -t x | tr -d ' ')
$ cat <<EOF > /etc/kubernetes/token.csv
${BOOTSTRAP_TOKEN},kubelet-bootstrap,10001,"system:kubelet-bootstrap"
EOF

# bootstrap set-cluster
$ kubectl config set-cluster kubernetes \
    --certificate-authority=ca.pem \
    --embed-certs=true \
    --server=${KUBE_APISERVER} \
    --kubeconfig=../bootstrap.conf

# bootstrap set-credentials
$ kubectl config set-credentials kubelet-bootstrap \
    --token=${BOOTSTRAP_TOKEN} \
    --kubeconfig=../bootstrap.conf

# bootstrap set-context
$ kubectl config set-context default \
    --cluster=kubernetes \
    --user=kubelet-bootstrap \
   --kubeconfig=../bootstrap.conf

# bootstrap set default context
$ kubectl config use-context default --kubeconfig=../bootstrap.conf

如果想用CA的方式来认证,可以参考Kubelet certificate

Admin certificate

下载admin-csr.json,并生成admin certificate证书:

$ wget "${PKI_URL}/admin-csr.json"
$ cfssl gencert \
  -ca=ca.pem \
  -ca-key=ca-key.pem \
  -config=ca-config.json \
  -profile=kubernetes \
  admin-csr.json | cfssljson -bare admin

$ ls admin*.pem
admin-key.pem  admin.pem  

然后执行一下命令生成名为admin.conf的kubeconfig:

# admin set-cluster
$ kubectl config set-cluster kubernetes \
    --certificate-authority=ca.pem \
    --embed-certs=true \
    --server=${KUBE_APISERVER} \
    --kubeconfig=../admin.conf

# admin set-credentials
$ kubectl config set-credentials kubernetes-admin \
    --client-certificate=admin.pem \
    --client-key=admin-key.pem \
    --embed-certs=true \
    --kubeconfig=../admin.conf

# admin set-context
$ kubectl config set-context kubernetes-admin@kubernetes \
    --cluster=kubernetes \
    --user=kubernetes-admin \
    --kubeconfig=../admin.conf

# admin set default context
$ kubectl config use-context kubernetes-admin@kubernetes \
    --kubeconfig=../admin.conf
Controller manager certificate

下载manager-csr.json,并生成kube-controller-manager certificate证书:

$ wget "${PKI_URL}/manager-csr.json"
$ cfssl gencert \
  -ca=ca.pem \
  -ca-key=ca-key.pem \
  -config=ca-config.json \
  -profile=kubernetes \
  manager-csr.json | cfssljson -bare controller-manager

$ ls controller-manager*.pem

如果节点IP不同,需要修改manager-csr.json的hosts

然后执行命令生成名为controller-manager.conf的kubeconfig:

# controller-manager set-cluster
$ kubectl config set-cluster kubernetes \
    --certificate-authority=ca.pem \
    --embed-certs=true \
    --server=${KUBE_APISERVER} \
    --kubeconfig=../controller-manager.conf

# controller-manager set-credentials
$ kubectl config set-credentials system:kube-controller-manager \
    --client-certificate=controller-manager.pem \
    --client-key=controller-manager-key.pem \
    --embed-certs=true \
    --kubeconfig=../controller-manager.conf

# controller-manager set-context
$ kubectl config set-context system:kube-controller-manager@kubernetes \
    --cluster=kubernetes \
    --user=system:kube-controller-manager \
    --kubeconfig=../controller-manager.conf

# controller-manager set default context
$ kubectl config use-context system:kube-controller-manager@kubernetes \
    --kubeconfig=../controller-manager.conf
Scheduler certificate

下载scheduler-csr.json,生成kube-scheduler certificate证书:

$ wget "${PKI_URL}/scheduler-csr.json"
$ cfssl gencert \
  -ca=ca.pem \
  -ca-key=ca-key.pem \
  -config=ca-config.json \
  -profile=kubernetes \
  scheduler-csr.json | cfssljson -bare scheduler

$ ls scheduler*.pem
scheduler-key.pem  scheduler.pem  

如果节点IP不同,需要修改scheduler-csr.json的hosts

然后执行一下命令生成名为scheduler.conf的kubeconfig:

# scheduler set-cluster
$ kubectl config set-cluster kubernetes \
    --certificate-authority=ca.pem \
    --embed-certs=true \
    --server=${KUBE_APISERVER} \
    --kubeconfig=../scheduler.conf

# scheduler set-credentials
$ kubectl config set-credentials system:kube-scheduler \
    --client-certificate=scheduler.pem \
    --client-key=scheduler-key.pem \
    --embed-certs=true \
    --kubeconfig=../scheduler.conf

# scheduler set-context
$ kubectl config set-context system:kube-scheduler@kubernetes \
    --cluster=kubernetes \
    --user=system:kube-scheduler \
    --kubeconfig=../scheduler.conf

# scheduler set default context
$ kubectl config use-context system:kube-scheduler@kubernetes \
    --kubeconfig=../scheduler.conf
Kubelet master certificate

下载kubelet-csr.json,并生成master node certificate证书:

$ wget "${PKI_URL}/kubelet-csr.json"
$ sed -i 's/$NODE/master1/g' kubelet-csr.json
$ cfssl gencert \
  -ca=ca.pem \
  -ca-key=ca-key.pem \
  -config=ca-config.json \
  -hostname=master1,172.16.35.12 \
  -profile=kubernetes \
  kubelet-csr.json | cfssljson -bare kubelet

$ ls kubelet*.pem
kubelet-key.pem  kubelet.pem  

$NODE需要随节点名称不同而改变

然后执行一下命令生成名为kubelet.conf的kubeconfig:

# kubelet set-cluster
$ kubectl config set-cluster kubernetes \
    --certificate-authority=ca.pem \
    --embed-certs=true \
    --server=${KUBE_APISERVER} \
    --kubeconfig=../kubelet.conf

# kubelet set-credentials
$ kubectl config set-credentials system:node:master1 \
    --client-certificate=kubelet.pem \
    --client-key=kubelet-key.pem \
    --embed-certs=true \
    --kubeconfig=../kubelet.conf

# kubelet set-context
$ kubectl config set-context system:node:master1@kubernetes \
    --cluster=kubernetes \
    --user=system:node:master1 \
    --kubeconfig=../kubelet.conf

# kubelet set default context
$ kubectl config use-context system:node:master1@kubernetes \
    --kubeconfig=../kubelet.conf
Service account key

Service account不需要CA认证,也就不需要CA来做Service account key的检查,这里我们建立一组private和public的密钥供Service account key使用:

$ openssl genrsa -out sa.key 2048 $ openssl rsa -in sa.key -pubout -out sa.pub $ ls sa.* sa.key sa.pub 

完成后删除不必要文件:

$ rm -rf *.json *.csr 

确认/etc/kubernetes/etc/kubernetes/pki包含以下文件:

$ ls /etc/kubernetes/
admin.conf  bootstrap.conf  controller-manager.conf  kubelet.conf  pki  scheduler.conf  token.csv

$ ls /etc/kubernetes/pki
admin-key.pem  apiserver-key.pem  ca-key.pem  controller-manager-key.pem  front-proxy-ca-key.pem  front-proxy-client-key.pem  kubelet-key.pem  sa.key  scheduler-key.pem  
admin.pem      apiserver.pem      ca.pem      controller-manager.pem      front-proxy-ca.pem      front-proxy-client.pem      kubelet.pem      sa.pub  scheduler.pem  

安装Kubernetes 核心元件

下载Kubernetes核心组件Yaml文件,这里我们利用Kubernetes Statics Pod来建立Master核心组件,因此下载所有Static Pod文件到etc/kubernetes/manifests目录“

$ export CORE_URL="https://kairen.github.io/files/manual-v1.8/master"
$ mkdir -p /etc/kubernetes/manifests && cd /etc/kubernetes/manifests
$ for FILE in apiserver manager scheduler; do
    wget "${CORE_URL}/${FILE}.yml.conf" -O ${FILE}.yml
  done

同样的,如果IP与本文IP准备不同的话,需要修改apiserver.ymlmanager.yml scheduler.yml

apiserver中的NodeRestriction请参考Using Node Authorization

生成一个用来加密Etcd的key

$ head -c 32 /dev/urandom | base64
SUpbL4juUYyvxj3/gonV5xVEx8j769/99TSAf8YT/sQ=  

/etc/kubernetes/目录建立encryption.yml的加密YAML文件:

$ cat <<EOF > /etc/kubernetes/encryption.yml
kind: EncryptionConfig  
apiVersion: v1  
resources:  
  - resources:
      - secrets
    providers:
      - aescbc:
          keys:
            - name: key1
              secret: SUpbL4juUYyvxj3/gonV5xVEx8j769/99TSAf8YT/sQ=
      - identity: {}
EOF  

Etcd加密可参考Encrypting data at rest

/etc/kubernetes/目录建立audit-policy.yml的auditing policay YAML文件:

$ cat <<EOF > /etc/kubernetes/audit-policy.yml
apiVersion: audit.k8s.io/v1beta1  
kind: Policy  
rules:  
- level: Metadata
EOF  

audit policy请参考Audit

下载kubelet.service相关文件来管理kubelet:

$ export KUBELET_URL="https://kairen.github.io/files/manual-v1.8/master"
$ mkdir -p /etc/systemd/system/kubelet.service.d
$ wget "${KUBELET_URL}/kubelet.service" -O /lib/systemd/system/kubelet.service
$ wget "${KUBELET_URL}/10-kubelet.conf" -O /etc/systemd/system/kubelet.service.d/10-kubelet.conf

cluster-dnscluster-domain有变动,需要修改10-kubelet.conf

最后建立var并启动kubelet服务:

$ mkdir -p /var/lib/kubelet /var/log/kubernetes
$ systemctl enable kubelet.service && systemctl start kubelet.service

完成后需要一段时间来下载镜像文件并启动组件:

$ watch netstat -ntlp
tcp        0      0 127.0.0.1:10248         0.0.0.0:*               LISTEN      23012/kubelet  
tcp        0      0 127.0.0.1:10251         0.0.0.0:*               LISTEN      22305/kube-schedule  
tcp        0      0 127.0.0.1:10252         0.0.0.0:*               LISTEN      22529/kube-controll  
tcp6       0      0 :::6443                 :::*                    LISTEN      22956/kube-apiserve  

看到上述信息即表明服务启动正常,如果出现问题可通过docker cli查看

完成后,复制admin kubeconfig并通过以下命令验证:

$ cp /etc/kubernetes/admin.conf ~/.kube/config
$ kubectl get cs
NAME                 STATUS    MESSAGE              ERROR  
etcd-0               Healthy   {"health": "true"}  
scheduler            Healthy   ok  
controller-manager   Healthy   ok

$ kubectl get node
NAME      STATUS     ROLES     AGE       VERSION  
master1   NotReady   master    1m        v1.8.6

$ kubectl -n kube-system get po
NAME                              READY     STATUS    RESTARTS   AGE  
kube-apiserver-master1            1/1       Running   0          4m  
kube-controller-manager-master1   1/1       Running   0          4m  
kube-scheduler-master1            1/1       Running   0          4m  

确认服务能够执行logs等命令:

$ kubectl -n kube-system logs -f kube-scheduler-master1
Error from server (Forbidden): Forbidden (user=kube-apiserver, verb=get, resource=nodes, subresource=proxy) ( pods/log kube-apiserver-master1)  

出现403 Forbidden问题表明kube-apiserver user并没有nodes的权限

由于上述权限问题,我们需要建立一个apiserver-to-kubelet-rbac.yml来定义权限,以供我们执行logs、exec等命令:

$ cd /etc/kubernetes/
$ export URL="https://kairen.github.io/files/manual-v1.8/master"
$ wget "${URL}/apiserver-to-kubelet-rbac.yml.conf" -O apiserver-to-kubelet-rbac.yml
$ kubectl apply -f apiserver-to-kubelet-rbac.yml

# 測試 logs
$ kubectl -n kube-system logs -f kube-scheduler-master1
...
I1031 03:22:42.527697       1 leaderelection.go:184] successfully acquired lease kube-system/kube-scheduler  

Kubernetes Node

Node运行容器实例的节点,即工作节点。本部分我们会下载Kubernetes binary并建立node 的certificate来提供给节点注册认证用。Kubernetes使用Node Authorizer来提供Authorization mode,这种授权模式会替Kubelet生成API request。

开始前,我们先在master1将需要的ca和cert复制到Node节点上:

$ for NODE in node1 node2; do
    ssh ${NODE} "mkdir -p /etc/kubernetes/pki/"
    ssh ${NODE} "mkdir -p /etc/etcd/ssl"
    # Etcd ca and cert
    for FILE in etcd-ca.pem etcd.pem etcd-key.pem; do
      scp /etc/etcd/ssl/${FILE} ${NODE}:/etc/etcd/ssl/${FILE}
    done
    # Kubernetes ca and cert
    for FILE in pki/ca.pem pki/ca-key.pem bootstrap.conf; do
      scp /etc/kubernetes/${FILE} ${NODE}:/etc/kubernetes/${FILE}
    done
  done

下载Kubernetes 元件

首先获取所有需要执行的文件:

# Download Kubernetes
$ export KUBE_URL="https://storage.googleapis.com/kubernetes-release/release/v1.8.6/bin/linux/amd64"
$ wget "${KUBE_URL}/kubelet" -O /usr/local/bin/kubelet
$ chmod +x /usr/local/bin/kubelet

# Download CNI
$ mkdir -p /opt/cni/bin && cd /opt/cni/bin
$ export CNI_URL="https://github.com/containernetworking/plugins/releases/download"
$ wget -qO- --show-progress "${CNI_URL}/v0.6.0/cni-plugins-amd64-v0.6.0.tgz" | tar -zx

设定Kubernetes node

下载Kubernetes相关文件,包括drop-in file、systemd service等:

$ export KUBELET_URL="https://kairen.github.io/files/manual-v1.8/node"
$ mkdir -p /etc/systemd/system/kubelet.service.d
$ wget "${KUBELET_URL}/kubelet.service" -O /lib/systemd/system/kubelet.service
$ wget "${KUBELET_URL}/10-kubelet.conf" -O /etc/systemd/system/kubelet.service.d/10-kubelet.conf

如果cluster-dnscluster-domain有改变的话,需要修改10-kubelet.conf

然后在所有node建立var,并启动kubelet服务:

$ mkdir -p /var/lib/kubelet /var/log/kubernetes /etc/kubernetes/manifests
$ systemctl enable kubelet.service && systemctl start kubelet.service

授权Kubernetes Node

重复完成所有节点后,在master1节点建立ClusterRoleBinding(因为我们采用的是TLS Bootstrapping):

$ kubectl create clusterrolebinding kubelet-bootstrap \
    --clusterrole=system:node-bootstrapper \
    --user=kubelet-bootstrap

在master进行验证,我们可以看到节点处于pending:

$ kubectl get csr
NAME                                                   AGE       REQUESTOR           CONDITION  
node-csr-YWf97ZrLCTlr2hmXsNLfjVLwaLfZRsu52FRKOYjpcBE   2s        kubelet-bootstrap   Pending  
node-csr-eq4q6ffOwT4yqYQNU6sT7mphPOQdFN6yulMVZeu6pkE   2s        kubelet-bootstrap   Pending  

通过kubectl,允许节点加入集群:

$ kubectl get csr | awk '/Pending/ {print $1}' | xargs kubectl certificate approve
certificatesigningrequest "node-csr-YWf97ZrLCTlr2hmXsNLfjVLwaLfZRsu52FRKOYjpcBE" approved  
certificatesigningrequest "node-csr-eq4q6ffOwT4yqYQNU6sT7mphPOQdFN6yulMVZeu6pkE" approved

$ kubectl get csr
NAME                                                   AGE       REQUESTOR           CONDITION  
node-csr-YWf97ZrLCTlr2hmXsNLfjVLwaLfZRsu52FRKOYjpcBE   30s       kubelet-bootstrap   Approved,Issued  
node-csr-eq4q6ffOwT4yqYQNU6sT7mphPOQdFN6yulMVZeu6pkE   30s       kubelet-bootstrap   Approved,Issued

$ kubectl get no
NAME      STATUS     ROLES     AGE       VERSION  
master1   NotReady   master    21m       v1.8.6  
node1     NotReady   node      8s        v1.8.6  
node2     NotReady   node      8s        v1.8.6  

Kubernetes Core Addons 部署

完成以上所有步骤,我们还需要安装一些插件,比如kube-dns、kube-proxy等等。

Kube-proxy addon

Kube-proxy是实现Service的关键组件,kube-proxy会在每个节点上执行,然后监听API Server的Service和Endpoint变化,并根据变化执行iptables实现网络转发。

这里我们需要DaemonSet来执行,并需要生成一些certificate。

首先在master1下载kube-proxy-csr.json,并生成kube-proxy certificate证书:

$ export PKI_URL="https://kairen.github.io/files/manual-v1.8/pki"
$ cd /etc/kubernetes/pki
$ wget "${PKI_URL}/kube-proxy-csr.json" "${PKI_URL}/ca-config.json"
$ cfssl gencert \
  -ca=ca.pem \
  -ca-key=ca-key.pem \
  -config=ca-config.json \
  -profile=kubernetes \
  kube-proxy-csr.json | cfssljson -bare kube-proxy

$ ls kube-proxy*.pem
kube-proxy-key.pem  kube-proxy.pem  

然后通过以下命令生成名为`kube-proxy.conf·的kubeconfig:

BOOTSTRAP_TOKEN

删除不必要的文件:

bootstrap.conf

确认/etc/kubernetes有以下文件:

$ export BOOTSTRAP_TOKEN=$(head -c 16 /dev/urandom | od -An -t x | tr -d ' ')
$ cat <<EOF > /etc/kubernetes/token.csv
${BOOTSTRAP_TOKEN},kubelet-bootstrap,10001,"system:kubelet-bootstrap"
EOF

# bootstrap set-cluster
$ kubectl config set-cluster kubernetes \
    --certificate-authority=ca.pem \
    --embed-certs=true \
    --server=${KUBE_APISERVER} \
    --kubeconfig=../bootstrap.conf

# bootstrap set-credentials
$ kubectl config set-credentials kubelet-bootstrap \
    --token=${BOOTSTRAP_TOKEN} \
    --kubeconfig=../bootstrap.conf

# bootstrap set-context
$ kubectl config set-context default \
    --cluster=kubernetes \
    --user=kubelet-bootstrap \
   --kubeconfig=../bootstrap.conf

# bootstrap set default context
$ kubectl config use-context default --kubeconfig=../bootstrap.conf

master1上将kube-proxy相关文件复制到Node节点上:

admin-csr.json

完成后,在master1通过kubectl建立kube-proxy daemon:

admin certificate

Kube-dns addon

Kube DNS是Kubernetes集群内部Pod之间通信的重要插件,允许Pod通过Domain Name链接Service,主要由Kube DNS与Sky DNS组合而成,通过Kube DNS监听Service与Endpoint变化,来提供给Sky DNS信息以更新解析位址。

安装只需要在master1通过kubectl建立kube-dns deployment即可:

$ wget "${PKI_URL}/admin-csr.json"
$ cfssl gencert \
  -ca=ca.pem \
  -ca-key=ca-key.pem \
  -config=ca-config.json \
  -profile=kubernetes \
  admin-csr.json | cfssljson -bare admin

$ ls admin*.pem
admin-key.pem  admin.pem  

Calico Network 安装与设定

Calico是一款纯3层协议(不需要Overlay 网路),已与各种云原生平台有良好的整合,在每个节点节点利用Linux Kernel实现高效的vRouter来负责数据转发,而当数据中心复杂度增加时,可以用BGP route reflector来达成。

首先在master1通过kubectl建立Calico policy controller:

admin.conf

如果节点IP不同,需要修改calico-controller.yml的ETCD_ENDPOINTS

在`master1·下载Calico CLI工具:

$ wget https://github.com/projectcalico/calicoctl/releases/download/v1.6.1/calicoctl
$ chmod +x calicoctl && mv calicoctl /usr/local/bin/

然后在所有节点下载Calico,并执行以下命令:

$ export CALICO_URL="https://github.com/projectcalico/cni-plugin/releases/download/v1.11.0"
$ wget -N -P /opt/cni/bin ${CALICO_URL}/calico
$ wget -N -P /opt/cni/bin ${CALICO_URL}/calico-ipam
$ chmod +x /opt/cni/bin/calico /opt/cni/bin/calico-ipam

接着在所有节点下载CNI plugins以及calico-node.service:

$ mkdir -p /etc/cni/net.d
$ export CALICO_CONF_URL="https://kairen.github.io/files/manual-v1.8/network"
$ wget "${CALICO_CONF_URL}/10-calico.conf" -O /etc/cni/net.d/10-calico.conf
$ wget "${CALICO_CONF_URL}/calico-node.service" -O /lib/systemd/system/calico-node.service

如果节点IP不同,需要修改10-calico.conf的etcd_endpoints

如果部署机器是虚拟机,需要修改calico-node.service,并在IPAUTODETECTIONMETHOD (包含IP6)部分指定绑定的网卡,以免预设绑定到NAT网路上

之后在所有节点启动Calico-node:

$ systemctl enable calico-node.service && systemctl start calico-node.service

master1查看Calico nodes:

$ cat <<EOF > ~/calico-rc
export ETCD_ENDPOINTS="https://172.16.35.12:2379"  
export ETCD_CA_CERT_FILE="/etc/etcd/ssl/etcd-ca.pem"  
export ETCD_CERT_FILE="/etc/etcd/ssl/etcd.pem"  
export ETCD_KEY_FILE="/etc/etcd/ssl/etcd-key.pem"  
EOF

$ . ~/calico-rc
$ calicoctl get node -o wide
NAME      ASN       IPV4              IPV6  
master1   (64512)   172.16.35.12/24  
node1     (64512)   172.16.35.10/24  
node2     (64512)   172.16.35.11/24  

查看pending的pod是否已执行:

$ kubectl -n kube-system get po
NAME                                        READY     STATUS    RESTARTS   AGE  
calico-policy-controller-5ff8b4549d-tctmm   1/1       Running   0          4m  
kube-apiserver-master1                      1/1       Running   0          20m  
kube-controller-manager-master1             1/1       Running   0          20m  
kube-dns-6cb549f55f-h4zr5                   3/3       Running   0          5m  
kube-proxy-fnrkb                            1/1       Running   0          6m  
kube-proxy-l72bq                            1/1       Running   0          6m  
kube-proxy-m6rfw                            1/1       Running   0          6m  
kube-scheduler-master1                      1/1       Running   0          20m  

省事的做法是用Standard Hosted方式安装。

Kubernetes Extra Addons 部署

本部分说明如何部署官方常用的addons,例如dashboard、heapster等。

Dashboard addon

Dashboard是Kubernetes官方开发的仪表板,让我们以可以i 通过web-based方式管理kubernetes集群。

master1通过kubectl建立kubernetes dashboard即可:

$ kubectl apply -f https://raw.githubusercontent.com/kubernetes/dashboard/master/src/deploy/recommended/kubernetes-dashboard.yaml
$ kubectl -n kube-system get po,svc -l k8s-app=kubernetes-dashboard
NAME                                      READY     STATUS    RESTARTS   AGE  
po/kubernetes-dashboard-747c4f7cf-md5m8   1/1       Running   0          56s

NAME                       TYPE        CLUSTER-IP      EXTERNAL-IP   PORT(S)   AGE  
svc/kubernetes-dashboard   ClusterIP   10.98.120.209   <none>        443/TCP   56s  

这边会额外建立一个名称为open-api的Cluster Role Binding,放拜年测试使用,一般情况下不开启(开启会存取所有API)。

$ cat <<EOF | kubectl create -f -
apiVersion: rbac.authorization.k8s.io/v1beta1  
kind: ClusterRoleBinding  
metadata:  
  name: open-api
  namespace: ""
roleRef:  
  apiGroup: rbac.authorization.k8s.io
  kind: ClusterRole
  name: cluster-admin
subjects:  
  - apiGroup: rbac.authorization.k8s.io
    kind: User
    name: system:anonymous
EOF  

管理者可以针对特定使用者来开放API存取权限,这里我们为了方便直接绑在cluster-admin cluster role。

1.7版本后的Dashboard不再提供所有权限,需要建立一个service account来绑定cluster-admin role:

$ kubectl -n kube-system create sa dashboard
$ kubectl create clusterrolebinding dashboard --clusterrole cluster-admin --serviceaccount=kube-system:dashboard
$ SECRET=$(kubectl -n kube-system get sa dashboard -o yaml | awk '/dashboard-token/ {print $3}')
$ kubectl -n kube-system describe secrets ${SECRET} | awk '/token:/{print $2}'
eyJhbGciOiJSUzI1NiIsInR5cCI6IkpXVCJ9.eyJpc3MiOiJrdWJlcm5ldGVzL3NlcnZpY2VhY2NvdW50Iiwia3ViZXJuZXRlcy5pby9zZXJ2aWNlYWNjb3VudC9uYW1lc3BhY2UiOiJrdWJlLXN5c3RlbSIsImt1YmVybmV0ZXMuaW8vc2VydmljZWFjY291bnQvc2VjcmV0Lm5hbWUiOiJkYXNoYm9hcmQtdG9rZW4tdzVocmgiLCJrdWJlcm5ldGVzLmlvL3NlcnZpY2VhY2NvdW50L3NlcnZpY2UtYWNjb3VudC5uYW1lIjoiZGFzaGJvYXJkIiwia3ViZXJuZXRlcy5pby9zZXJ2aWNlYWNjb3VudC9zZXJ2aWNlLWFjY291bnQudWlkIjoiYWJmMTFjYzMtZjRlYi0xMWU3LTgzYWUtMDgwMDI3NjdkOWI5Iiwic3ViIjoic3lzdGVtOnNlcnZpY2VhY2NvdW50Omt1YmUtc3lzdGVtOmRhc2hib2FyZCJ9.Xuyq34ci7Mk8bI97o4IldDyKySOOqRXRsxVWIJkPNiVUxKT4wpQZtikNJe2mfUBBD-JvoXTzwqyeSSTsAy2CiKQhekW8QgPLYelkBPBibySjBhJpiCD38J1u7yru4P0Pww2ZQJDjIxY4vqT46ywBklReGVqY3ogtUQg-eXueBmz-o7lJYMjw8L14692OJuhBjzTRSaKW8U2MPluBVnD7M2SOekDff7KpSxgOwXHsLVQoMrVNbspUCvtIiEI1EiXkyCNRGwfnd2my3uzUABIHFhm0_RZSmGwExPbxflr8Fc6bxmuz-_jSdOtUidYkFIzvEWw2vRovPgs3MXTv59RwUw  

复制token,然后贴到Kubernetes dashboard

Heapster addon

Heapster是Kubernetes社区维护的容器集群监控和分析工具。Heapster会从Kubernetes apiserver取得所有Node数据,然后再通过Node获取kubelet上的数据,最后再将所有收集到数据送到Heapster后台储存InfluxDB,最后利用Grafana抓取InfluxDB数据源来进行展示。

master1通过kubectl来建立kubernetes monitor即可:

$ export ADDON_URL="https://kairen.github.io/files/manual-v1.8/addon"
$ wget ${ADDON_URL}/kube-monitor.yml.conf -O kube-monitor.yml
$ kubectl apply -f kube-monitor.yml
$ kubectl -n kube-system get po,svc
NAME                                           READY     STATUS    RESTARTS   AGE  
...
po/heapster-74fb5c8cdc-62xzc                   4/4       Running   0          7m  
po/influxdb-grafana-55bd7df44-nw4nc            2/2       Running   0          7m

NAME                       TYPE        CLUSTER-IP       EXTERNAL-IP   PORT(S)             AGE  
...
svc/heapster               ClusterIP   10.100.242.225   <none>        80/TCP              7m  
svc/monitoring-grafana     ClusterIP   10.101.106.180   <none>        80/TCP              7m  
svc/monitoring-influxdb    ClusterIP   10.109.245.142   <none>        8083/TCP,8086/TCP   7m  
···

简单部署Nginx 服务

Kubernetes可以选择使用指令直接建立应用和服务,或者我们可以写YAML、JSON文件来配置,如下所示:

$ kubectl run nginx --image=nginx --port=80
$ kubectl expose deploy nginx --port=80 --type=LoadBalancer --external-ip=172.16.35.12
$ kubectl get svc,po
NAME             TYPE           CLUSTER-IP      EXTERNAL-IP    PORT(S)        AGE  
svc/kubernetes   ClusterIP      10.96.0.1       <none>         443/TCP        1h  
svc/nginx        LoadBalancer   10.97.121.243   172.16.35.12   80:30344/TCP   22s

NAME                        READY     STATUS    RESTARTS   AGE  
po/nginx-7cbc4b4d9c-7796l   1/1       Running   0          28s       192.160.57.181   ,172.16.35.12   80:32054/TCP   21s  

这里type可以选择NodePort和LoadBalancer在本地裸机部署,两者差异在于NodePort只映射Host port到Container port,而LoadBalancer则继承NodePort额外映射Host target port到Container port

扩展服务数量

最后,我们可以通过以下方式来扩展服务数量:

$ kubectl scale deploy nginx --replicas=2

$ kubectl get pods -o wide
NAME                    READY     STATUS    RESTARTS   AGE       IP             NODE  
nginx-158599303-0h9lr   1/1       Running   0          25s       10.244.100.5   node2  
nginx-158599303-k7cbt   1/1       Running   0          1m        10.244.24.3    node1  

相关阅读

开源PaaS Rainbond提供生产就绪的Kubernetes,在线体验请注册公有云(新用户7天免费)