01 搭建K8s集群
官网
:https://kubernetes.io/docs/setup/production-environment/tools/kubeadm/install-kubeadm/#installing-kubeadm-kubelet-and-kubectl
GitHub
:https://github.com/kubernetes/kubeadm
课程中
:使用kubeadm搭建一个3台机器组成的k8s集群,1台master节点,2台worker节点如果大家机器配置不够,也可以使用在线的,或者minikube的方式或者1个master和1个worker
配置要求
:
- One or more machines running one of:
- Ubuntu 16.04+
- Debian 9+
- CentOS 7
- Red Hat Enterprise Linux (RHEL) 7
- Fedora 25+
- HypriotOS v1.0.1+
- Container Linux (tested with 1800.6.0)
- 2 GB or more of RAM per machine (any less will leave little room for your apps)
- 2 CPUs or more
- Full network connectivity between all machines in the cluster (public or private network is fine)
- Unique hostname, MAC address, and product_uuid for every node. See here for more details.
- Certain ports are open on your machines. See here for more details.
- Swap disabled. You MUST disable swap in order for the kubelet to work properly.
1.1 版本统一
Docker 18.09.0
---
kubeadm-1.14.0-0
kubelet-1.14.0-0
kubectl-1.14.0-0
---
k8s.gcr.io/kube-apiserver:v1.14.0
k8s.gcr.io/kube-controller-manager:v1.14.0
k8s.gcr.io/kube-scheduler:v1.14.0
k8s.gcr.io/kube-proxy:v1.14.0
k8s.gcr.io/pause:3.1
k8s.gcr.io/etcd:3.3.10
k8s.gcr.io/coredns:1.3.1
---
calico:v3.9
1.2 准备3台centos
大家根据自己的情况来准备centos7的虚拟机。
要保证彼此之间能够ping通,也就是处于同一个网络中,虚拟机的配置要求上面也描述咯。
1.3 更新并安装依赖
3台机器都需要执行
yum -y update
yum install -y conntrack ipvsadm ipset jq sysstat curl iptables libseccomp
1.4 安装Docker
根据之前学习的Docker方式[Docker第一节课的笔记中也有这块的说明]
在每一台机器上都安装好Docker,版本为18.09.0
01 安装必要的依赖 sudo yum install -y yum-utils \ device-mapper-persistent-data \ lvm2 02 设置docker仓库 sudo yum-config-manager --add-repo http://mirrors.aliyun.com/docker-ce/linux/centos/docker-ce.repo 【设置要设置一下阿里云镜像加速器】 sudo mkdir -p /etc/docker sudo tee /etc/docker/daemon.json <<-'EOF' { "registry-mirrors": ["这边替换成自己的实际地址"] } EOF sudo systemctl daemon-reload 03 安装docker yum install -y docker-ce-18.09.0 docker-ce-cli-18.09.0 containerd.io 04 启动docker sudo systemctl start docker && sudo systemctl enable docker
1.5 修改hosts文件
(1)master
# 设置master的hostname,并且修改hosts文件
sudo hostnamectl set-hostname m
vi /etc/hosts
192.168.8.51 m
192.168.8.61 w1
192.168.8.62 w2
(2)两个worker
# 设置worker01/02的hostname,并且修改hosts文件
sudo hostnamectl set-hostname w1
sudo hostnamectl set-hostname w2
vi /etc/hosts
192.168.8.51 m
192.168.8.61 w1
192.168.8.62 w2
(3)使用ping测试一下
1.6 系统基础前提配置
# (1)关闭防火墙
systemctl stop firewalld && systemctl disable firewalld
# (2)关闭selinux
setenforce 0
sed -i 's/^SELINUX=enforcing$/SELINUX=permissive/' /etc/selinux/config
# (3)关闭swap
swapoff -a
sed -i '/swap/s/^\(.*\)$/#\1/g' /etc/fstab
# (4)配置iptables的ACCEPT规则
iptables -F && iptables -X && iptables -F -t nat && iptables -X -t nat && iptables -P FORWARD ACCEPT
# (5)设置系统参数
cat < /etc/sysctl.d/k8s.conf
net.bridge.bridge-nf-call-ip6tables = 1
net.bridge.bridge-nf-call-iptables = 1
EOF
sysctl --system
1.7 Installing kubeadm, kubelet and kubectl
(1)配置yum源
cat < /etc/yum.repos.d/kubernetes.repo
[kubernetes]
name=Kubernetes
baseurl=http://mirrors.aliyun.com/kubernetes/yum/repos/kubernetes-el7-x86_64
enabled=1
gpgcheck=0
repo_gpgcheck=0
gpgkey=http://mirrors.aliyun.com/kubernetes/yum/doc/yum-key.gpg
http://mirrors.aliyun.com/kubernetes/yum/doc/rpm-package-key.gpg
EOF
(2)安装kubeadm&kubelet&kubectl
yum install -y kubeadm-1.14.0-0 kubelet-1.14.0-0 kubectl-1.14.0-0
(3)docker和k8s设置同一个cgroup
# docker
vi /etc/docker/daemon.json
"exec-opts": ["native.cgroupdriver=systemd"],
systemctl restart docker
# kubelet,这边如果发现输出directory not exist,也说明是没问题的,大家继续往下进行即可
sed -i "s/cgroup-driver=systemd/cgroup-driver=cgroupfs/g" /etc/systemd/system/kubelet.service.d/10-kubeadm.conf
systemctl enable kubelet && systemctl start kubelet
1.8 proxy/pause/scheduler等国内镜像
(1)查看kubeadm使用的镜像
kubeadm config images list
可以发现这里都是国外的镜像
k8s.gcr.io/kube-apiserver:v1.14.0
k8s.gcr.io/kube-controller-manager:v1.14.0
k8s.gcr.io/kube-scheduler:v1.14.0
k8s.gcr.io/kube-proxy:v1.14.0
k8s.gcr.io/pause:3.1
k8s.gcr.io/etcd:3.3.10
k8s.gcr.io/coredns:1.3.1
(2)解决国外镜像不能访问的问题
- 创建kubeadm.sh脚本,用于拉取镜像/打tag/删除原有镜像
#!/bin/bash
set -e
KUBE_VERSION=v1.14.0
KUBE_PAUSE_VERSION=3.1
ETCD_VERSION=3.3.10
CORE_DNS_VERSION=1.3.1
GCR_URL=k8s.gcr.io
ALIYUN_URL=registry.cn-hangzhou.aliyuncs.com/google_containers
images=(kube-proxy:${KUBE_VERSION}
kube-scheduler:${KUBE_VERSION}
kube-controller-manager:${KUBE_VERSION}
kube-apiserver:${KUBE_VERSION}
pause:${KUBE_PAUSE_VERSION}
etcd:${ETCD_VERSION}
coredns:${CORE_DNS_VERSION})
for imageName in ${images[@]} ; do
docker pull $ALIYUN_URL/$imageName
docker tag $ALIYUN_URL/$imageName $GCR_URL/$imageName
docker rmi $ALIYUN_URL/$imageName
done
- 运行脚本和查看镜像
# 运行脚本
sh ./kubeadm.sh
# 查看镜像
docker images
- 将这些镜像推送到自己的阿里云仓库【可选,根据自己实际的情况】
# 登录自己的阿里云仓库
docker login --username=xxx registry.cn-hangzhou.aliyuncs.com
#!/bin/bash
set -e
KUBE_VERSION=v1.14.0
KUBE_PAUSE_VERSION=3.1
ETCD_VERSION=3.3.10
CORE_DNS_VERSION=1.3.1
GCR_URL=k8s.gcr.io
ALIYUN_URL=xxx
images=(kube-proxy:${KUBE_VERSION}
kube-scheduler:${KUBE_VERSION}
kube-controller-manager:${KUBE_VERSION}
kube-apiserver:${KUBE_VERSION}
pause:${KUBE_PAUSE_VERSION}
etcd:${ETCD_VERSION}
coredns:${CORE_DNS_VERSION})
for imageName in ${images[@]} ; do
docker tag $GCR_URL/$imageName $ALIYUN_URL/$imageName
docker push $ALIYUN_URL/$imageName
docker rmi $ALIYUN_URL/$imageName
done
运行脚本 sh ./kubeadm-push-aliyun.sh
1.9 kube init初始化master
(1)kube init流程
01-进行一系列检查,以确定这台机器可以部署kubernetes
02-生成kubernetes对外提供服务所需要的各种证书可对应目录
/etc/kubernetes/pki/*
03-为其他组件生成访问kube-ApiServer所需的配置文件
ls /etc/kubernetes/
admin.conf controller-manager.conf kubelet.conf scheduler.conf
04-为 Master组件生成Pod配置文件。
ls /etc/kubernetes/manifests/*.yaml
kube-apiserver.yaml
kube-controller-manager.yaml
kube-scheduler.yaml
05-生成etcd的Pod YAML文件。
ls /etc/kubernetes/manifests/*.yaml
kube-apiserver.yaml
kube-controller-manager.yaml
kube-scheduler.yaml
etcd.yaml
06-一旦这些 YAML 文件出现在被 kubelet 监视的/etc/kubernetes/manifests/目录下,kubelet就会自动创建这些yaml文件定义的pod,即master组件的容器。master容器启动后,kubeadm会通过检查localhost:6443/healthz这个master组件的健康状态检查URL,等待master组件完全运行起来
07-为集群生成一个bootstrap token
08-将ca.crt等 Master节点的重要信息,通过ConfigMap的方式保存在etcd中,工后续部署node节点使用
09-最后一步是安装默认插件,kubernetes默认kube-proxy和DNS两个插件是必须安装的
(2)初始化master节点
官网:https://kubernetes.io/docs/reference/setup-tools/kubeadm/kubeadm/
注意
:此操作是在主节点上进行
# 本地有镜像
kubeadm init --kubernetes-version=1.14.0 --apiserver-advertise-address=192.168.8.51 --pod-network-cidr=10.244.0.0/16
【若要重新初始化集群状态:kubeadm reset,然后再进行上述操作】
记得保存好最后kubeadm join的信息
(3)根据日志提示
mkdir -p $HOME/.kube
sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
sudo chown $(id -u):$(id -g) $HOME/.kube/config
此时kubectl cluster-info查看一下是否成功
(4)查看pod验证一下
等待一会儿,同时可以发现像etc,controller,scheduler等组件都以pod的方式安装成功了
注意
:coredns没有启动,需要安装网络插件
kubectl get pods -n kube-system
(5)健康检查
curl -k https://localhost:6443/healthz
1.10 部署calico网络插件
选择网络插件:https://kubernetes.io/docs/concepts/cluster-administration/addons/
calico网络插件:https://docs.projectcalico.org/v3.9/getting-started/kubernetes/
calico,同样在master节点上操作
# 在k8s中安装calico
kubectl apply -f https://docs.projectcalico.org/v3.9/manifests/calico.yaml
# 确认一下calico是否安装成功
kubectl get pods --all-namespaces -w
1.11 kube join
记得保存初始化master节点的最后打印信息【注意这边大家要自己的,下面我的只是一个参考】
kubeadm join 192.168.0.51:6443 --token yu1ak0.2dcecvmpozsy8loh \
--discovery-token-ca-cert-hash sha256:5c4a69b3bb05b81b675db5559b0e4d7972f1d0a61195f217161522f464c307b0
(1)在woker01和worker02上执行上述命令
(2)在master节点上检查集群信息
kubectl get nodes
NAME STATUS ROLES AGE VERSION
master-kubeadm-k8s Ready master 19m v1.14.0
worker01-kubeadm-k8s Ready 3m6s v1.14.0
worker02-kubeadm-k8s Ready 2m41s v1.14.0
1.12 再次体验Pod
(1)定义pod.yml文件,比如pod_nginx_rs.yaml
cat > pod_nginx_rs.yaml <
(2)根据pod_nginx_rs.yml文件创建pod
kubectl apply -f pod_nginx_rs.yaml
(3)查看pod
kubectl get pods
kubectl get pods -o wide
kubectl describe pod nginx
(4)感受通过rs将pod扩容
kubectl scale rs nginx --replicas=5
kubectl get pods -o wide
(5)删除pod
kubectl delete -f pod_nginx_rs.yaml
02 Basic
2.1 yaml文件
2.1.1 简介
YAML(IPA: /ˈjæməl/)是一个可读性高的语言,参考了XML、C、Python等。
理解:Yet Another Markup Language
后缀:可以是.yml或者是.yaml,更加推荐.yaml,其实用任意后缀都可以,只是阅读性不强
2.1.2 基础
- 区分大小写
- 缩进表示层级关系,相同层级的元素左对齐
- 缩进只能使用空格,不能使用TAB
- "#"表示当前行的注释
- 是JSON文件的超级,两个可以转换
- ---表示分隔符,可以在一个文件中定义多个结构
- 使用key: value,其中":"和value之间要有一个英文空格
2.1.3 Maps
2.1.3.1 简单
apiVersion: v1
kind: Pod
---表示分隔符,可选。要定义多个结构一定要分隔
apiVersion表示key,v1表示value,英文":"后面要有一个空格
kind表示key,Pod表示value
也可以这样写apiVersion: "v1"
转换为JSON格式
{ "apiVersion": "v1", "kind": "Pod" }
apiVersion: apps/v1
kind: Deployment
metadata:
name: nginx-deployment
labels:
app: nginx
2.1.2.2 复杂
apiVersion: apps/v1
kind: Deployment
metadata:
name: nginx-deployment
labels:
app: nginx
metadata表示key,下面的内容表示value,该value中包含两个直接的key:name和labels
name表示key,nginx-deployment表示value
labels表示key,下面的表示value,这个值又是一个map
app表示key,nginx表示value
相同层级的记得使用空间缩进,左对齐
转换为JSON格式
{ "apiVersion": "apps/v1", "kind": "Deployment", "metadata": { "name": "nginx-deployment", "labels": { "app": "nginx" } } }
2.1.4 Lists
apiVersion: v1
kind: Pod
metadata:
name: myapp-pod
labels:
app: myapp
spec:
containers:
- name: myapp-container01
image: busybox:1.28
- name: myapp-container02
image: busybox:1.28
containers表示key,下面的表示value,其中value是一个数组
数组中有两个元素,每个元素里面包含name和image
image表示key,myapp-container表示value
转换成JSON格式
{ "apiVersion": "v1", "kind": "Pod", "metadata": { "name": "myapp", "labels": { "app": "myapp" } }, "spec": { "containers": [{ "name": "myapp-container01", "image": "busybox:1.28", }, { "name": "myapp-container02", "image": "busybox:1.28", }] } }
2.1.5 找个k8s的yaml文件
官网
:https://kubernetes.io/docs/reference/
# yaml格式对于Pod的定义:
apiVersion: v1 #必写,版本号,比如v1
kind: Pod #必写,类型,比如Pod
metadata: #必写,元数据
name: nginx #必写,表示pod名称
namespace: default #表示pod名称属于的命名空间
labels:
app: nginx #自定义标签名字
spec: #必写,pod中容器的详细定义
containers: #必写,pod中容器列表
- name: nginx #必写,容器名称
image: nginx #必写,容器的镜像名称
ports:
- containerPort: 80 #表示容器的端口
2.2 Container
官网
:https://kubernetes.io/docs/concepts/containers/
2.2.1 Docker世界中
可以通过docker run运行一个容器
或者定义一个yml文件,本机使用docker-compose,多机通过docker swarm创建
2.2.2 K8S世界中
同样以一个yaml文件维护,container运行在pod中
2.3 Pod
官网
:https://kubernetes.io/docs/concepts/workloads/pods/pod-overview/
2.3.1 What is Pod
A Pod is the basic execution unit of a Kubernetes application
A Pod encapsulates an application’s container (or, in some cases, multiple containers), storage resources, a unique network IP, and options that govern how the container(s) should run
2.3.2 Pod初体验
(1)创建一个pod的yaml文件,名称为nginx_pod.yaml
apiVersion: v1
kind: Pod
metadata:
name: nginx-pod
labels:
app: nginx
spec:
containers:
- name: nginx-container
image: nginx
ports:
- containerPort: 80
(2)根据该nginx_pod.yaml文件创建pod
kubectl apply -f nginx_pod.yaml
(3)查看pod
01 kubectl get pods
NAME READY STATUS RESTARTS AGE
nginx-pod 1/1 Running 0 29s
02 kubectl get pods -o wide
NAME READY STATUS RESTARTS AGE IP NODE
nginx-pod 1/1 Running 0 40m 192.168.80.194 w2
03 kubectl describe pod nginx-pod
Name: nginx-pod
Namespace: default
Priority: 0
PriorityClassName:
Node: w2/192.168.0.62
Start Time: Sun, 06 Oct 2019 20:45:35 +0000
Labels: app=nginx
Annotations: cni.projectcalico.org/podIP: 192.168.80.194/32
kubectl.kubernetes.io/last-applied-configuration:
{"apiVersion":"v1","kind":"Pod","metadata":{"annotations":{},"labels":{"app":"nginx"},"name":"nginx-pod","namespace":"default"},"spec":{"c...
Status: Running
IP: 192.168.80.194
Containers:
nginx-container:
Container ID: docker://eb2fd0b2906f53e9892e22a6fd791c9ac68fb8e5efce3bbf94ec12bae96e1984
Image: nginx
Image ID: docker-pullable:/
(4)可以发现该pod运行在worker02节点上
于是来到worker02节点,docker ps一下
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
eb2fd0b2906f nginx "nginx -g 'daemon of…" 6 minutes ago Up 6 minutes k8s_nginx-container_nginx-pod_default_3ee0706d-e87a-11e9-a904-5254008afee6_0
不妨进入该容器试试[可以发现只有在worker02上有该容器,因为pod运行在worker02上]:
docker exec -it k8s_nginx-container_nginx-pod_default_3ee0706d-e87a-11e9-a904-5254008afee6_0 bash
root@nginx-pod:/#
(5)访问nginx容器
curl 192.168.80.194 OK,并且在任何一个集群中的Node上访问都成功
(6)删除Pod
kubectl delete -f nginx_pod.yaml
kubectl get pods
2.3.3 Storage and Networking
官网
:https://kubernetes.io/docs/concepts/workloads/pods/pod-overview/#networking
- Networking
Each Pod is assigned a unique IP address. Every container in a Pod shares the network namespace, including the IP address and network ports.
官网
:https://kubernetes.io/docs/concepts/workloads/pods/pod-overview/#storage
- Storage
A Pod can specify a set of shared storage Volumes. All containers in the Pod can access the shared volumes, allowing those containers to share data.