K8S-Rails.md

条件

# 安装时需要确保 K8S 集群正在运行，并且挂载了 GlusterFS 集群文件系统

# 相关文档： K8S 集群安装配置.md、GlusterFS 集群安装配置.md

# 在各个主机检查 GlusterFS 集群文件系统

df -h /mnt/gluster-gv0

# 正常输出

Filesystem      Size  Used Avail Use% Mounted on
gfs:/gv0         48T  835G   48T   2% /mnt/gluster-gv0

以下操作均在 K8S 第一个节点，即 Master 节点 192.168.1.21 执行

准备

# 准备 Rails 应用，以 Winter-rails 为例，复制到 /mnt/gluster-gv0/k8s/rails 目录下

# 如果访问 GitHub.com 需要代理在此设置
# export http_proxy=http://192.168.1.109:3128/
# export https_proxy=http://192.168.1.109:3128/

(
	if [ ! -d ~/Winter-rails ]; then cd ~/ ; git clone https://github.com/AndyInAi/Winter-rails.git; fi ;

	mkdir -p /mnt/gluster-gv0/k8s/rails ;

	cp -f -r ~/Winter-rails /mnt/gluster-gv0/k8s/rails/ ;
)

配置

# 生成启动脚本

(
	mkdir -p /mnt/gluster-gv0/k8s/rails

	echo '#!/bin/bash

		echo "
		  deb http://mirrors.tuna.tsinghua.edu.cn/ubuntu/ jammy main restricted universe multiverse
		  deb http://mirrors.tuna.tsinghua.edu.cn/ubuntu/ jammy-updates main restricted universe multiverse
		  deb http://mirrors.tuna.tsinghua.edu.cn/ubuntu/ jammy-backports main restricted universe multiverse
		  deb http://mirrors.tuna.tsinghua.edu.cn/ubuntu/ jammy-security main restricted universe multiverse
		" > /etc/apt/sources.list

		export DEBIAN_FRONTEND=noninteractive

		apt update -y

		apt install -y rails ruby-xpath libyaml-dev ruby-tzinfo

		cp -f -r /mnt/rails/Winter-rails ~/

		cd  ~/Winter-rails

		gem sources --add http://mirrors.tuna.tsinghua.edu.cn/rubygems/ --remove https://rubygems.org/

		gem install xpath

		bundle config mirror.https://rubygems.org https://gems.ruby-china.com

		bundle install

		chmod +x bin/rails

		bin/rails server -b 0.0.0.0 -p 8080

		# sleep 88888

	'> /mnt/gluster-gv0/k8s/rails/start 
	
	chmod +x /mnt/gluster-gv0/k8s/rails/start
)

部署

# 副本：8 个

# 端口：30011

(
	mkdir -p /mnt/gluster-gv0/k8s/rails;

	kubectl get namespace winter > /dev/null 2>&1 || kubectl create namespace winter;

	echo '
                    apiVersion: apps/v1
                    kind: Deployment
                    metadata:
                      labels:
                        app: rails
                      name: rails
                      namespace: winter
                    spec:
                      replicas: 8
                      selector:
                        matchLabels:
                          app: rails
                      template:
                        metadata:
                          labels:
                            app: rails
                        spec:
                          volumes:
                            - name: k8s
                              hostPath:
                                path: /mnt/gluster-gv0/k8s
                          containers:
                            - image: ubuntu
                              imagePullPolicy: IfNotPresent
                              name: rails
                              volumeMounts:
                                - name: k8s
                                  mountPath: /mnt
                              command: ["/mnt/rails/start"]
	' > ~/rails.yml;

	echo '
		apiVersion: v1
		kind: Service
		metadata:
		  name: rails
		  namespace: winter
		spec:
		  selector:
		    app: rails
		  type: LoadBalancer
		  ports:
		    - nodePort: 30011
		      port: 8080
		      protocol: TCP
		      targetPort: 8080
		  sessionAffinity: ClientIP
	' > rails-port.yml;
	sed -i 's/\t/        /g' ~/rails.yml;
	sed -i 's/\t/        /g' ~/rails-port.yml;
	
	kubectl apply -f ~/rails.yml;
	kubectl apply -f ~/rails-port.yml;
)

查看

kubectl get pod,svc,node -n winter -o wide

测试

curl http://192.168.1.22:30011/
curl http://192.168.1.23:30011/
curl http://192.168.1.24:30011/

# 使用浏览器打开

http://192.168.1.24:30011/
http://192.168.1.23:30011/
http://192.168.1.22:30011/

扩缩

# 静态扩缩，指定 pod 数量，如 12

kubectl scale --replicas=12 -f ~/rails.yml

# 动态扩缩，指定最小和最大 pod 数量

kubectl autoscale  --max=16 --min=4 -f ~/rails.yml