春风十里不如你 —— Taozi - 集群

基于Kubernetes集群的监控网络服务

2023-10-30T01:41:00+08:00

基于Kubernetes集群的监控网络服务介绍需要以下环境Kubernetes集群Blackbox工具Grafana、Prometheus监控大致功能：通过在K8s集群中部署blackbox工具（用于监控服务，检查网络可用性）和Grafana、Prometheus（监控可视化面板）更直观的体现网络连通性，可以进行警报和分析本文章通过若海博客的【Kubernetes 集群上安装 Blackbox 监控网站状态】和【Kubernetes 集群上安装 Grafana 和 Prometheus】整合而成部署Kubernetes集群（Ubuntu/Debian操作系统）确保主节点和子节点都有Docker环境（最好是同一个版本）主节点//安装Docker，一键安装（如有安装可以忽略） curl -fsSL https://get.docker.com | bash -s docker --mirror Aliyun //开启docker、并设置开机自启 systemctl start docker & systemctl enable docker apt update apt install -y wireguard echo "net.ipv4.ip_forward = 1" >/etc/sysctl.d/ip_forward.conf sysctl -p /etc/sysctl.d/ip_forward.conf //以下Token值请保存，任意字符串 export SERVER_TOKEN=r83nui54eg8wihyiteshuo3o43gbf7u9er63o43gbf7uitujg8wihyitr6 export PUBLIC_IP=$(curl -Ls http://metadata.tencentyun.com/latest/meta-data/public-ipv4) export PRIVATE_IP=$(curl -Ls http://metadata.tencentyun.com/latest/meta-data/local-ipv4) export INSTALL_K3S_SKIP_DOWNLOAD=true export DOWNLOAD_K3S_BIN_URL=https://github.com/k3s-io/k3s/releases/download/v1.28.2%2Bk3s1/k3s if [ $(curl -Ls http://ipip.rehi.org/country_code) == "CN" ]; then DOWNLOAD_K3S_BIN_URL=https://ghproxy.com/${DOWNLOAD_K3S_BIN_URL} fi curl -Lo /usr/local/bin/k3s $DOWNLOAD_K3S_BIN_URL chmod a+x /usr/local/bin/k3s curl -Ls https://get.k3s.io | sh -s - server \ --cluster-init \ --token $SERVER_TOKEN \ --node-ip $PRIVATE_IP \ --node-external-ip $PUBLIC_IP \ --advertise-address $PRIVATE_IP \ --service-node-port-range 5432-9876 \ --flannel-backend wireguard-native \ --flannel-external-ip子节点//安装Docker，一键安装（如有安装可以忽略） curl -fsSL https://get.docker.com | bash -s docker --mirror Aliyun //开启docker、并设置开机自启 systemctl start docker & systemctl enable docker //子节点代码 apt update apt install -y wireguard echo "net.ipv4.ip_forward = 1" >/etc/sysctl.d/ip_forward.conf sysctl -p /etc/sysctl.d/ip_forward.conf export SERVER_IP=43.129.195.33 //此ip填你的主节点地址 export SERVER_TOKEN=r83nui54eg8wihyiteshuo3o43gbf7u9er63o43gbf7uitujg8wihyitr6 export PUBLIC_IP=$(curl -Ls http://metadata.tencentyun.com/latest/meta-data/public-ipv4) export PRIVATE_IP=$(curl -Ls http://metadata.tencentyun.com/latest/meta-data/local-ipv4) export INSTALL_K3S_SKIP_DOWNLOAD=true export DOWNLOAD_K3S_BIN_URL=https://github.com/k3s-io/k3s/releases/download/v1.28.2%2Bk3s1/k3s if [ $(curl -Ls http://ipip.rehi.org/country_code) == "CN" ]; then DOWNLOAD_K3S_BIN_URL=https://ghproxy.com/${DOWNLOAD_K3S_BIN_URL} fi curl -Lo /usr/local/bin/k3s $DOWNLOAD_K3S_BIN_URL chmod a+x /usr/local/bin/k3s curl -Ls https://get.k3s.io | sh -s - agent \ --server https://$SERVER_IP:6443 \ --token $SERVER_TOKEN \ --node-ip $PRIVATE_IP \ --node-external-ip $PUBLIC_IPBlackbox工具部署（也有集群方式）//拉取镜像 docker pull rehiy/blackbox //一键启动 docker run -d \ --name blackbox \ --restart always \ --publish 9115:9115 \ --env "NODE_NAME=guangzhou-taozi" \ --env "NODE_OWNER=Taozi" \ --env "NODE_REGION=广州" \ --env "NODE_ISP=TencentCloud" \ --env "NODE_BANNER=From Taozii-www.xiongan.host" \ rehiy/blackbox //开始注册 docker logs -f blackboxGrafana、Prometheus部署在主节点创建一个目录，名字任意，然后在同一目录中创建两个文件（grafpro.yaml、grafpro.sh）grafpro.yamlkind: Deployment apiVersion: apps/v1 metadata: name: &name grafpro labels: app: *name spec: selector: matchLabels: app: *name template: metadata: labels: app: *name spec: initContainers: - name: busybox image: busybox command: - sh - -c - | if [ ! -f /etc/prometheus/prometheus.yml ]; then cat </etc/prometheus/prometheus.yml global: scrape_timeout: 25s scrape_interval: 1m evaluation_interval: 1m scrape_configs: - job_name: prometheus static_configs: - targets: - 127.0.0.1:9090 EOF fi volumeMounts: - name: *name subPath: etc mountPath: /etc/prometheus containers: - name: grafana image: grafana/grafana securityContext: runAsUser: 0 ports: - containerPort: 3000 volumeMounts: - name: *name subPath: grafana mountPath: /var/lib/grafana - name: prometheus image: prom/prometheus securityContext: runAsUser: 0 ports: - containerPort: 9090 volumeMounts: - name: *name subPath: etc mountPath: /etc/prometheus - name: *name subPath: prometheus mountPath: /prometheus volumes: - name: *name hostPath: path: /srv/grafpro type: DirectoryOrCreate --- kind: Service apiVersion: v1 metadata: name: &name grafpro labels: app: *name spec: selector: app: *name ports: - name: grafana port: 3000 targetPort: 3000 - name: prometheus port: 9090 targetPort: 9090 --- kind: Ingress apiVersion: networking.k8s.io/v1 metadata: name: &name grafpro annotations: traefik.ingress.kubernetes.io/router.entrypoints: web,websecure spec: rules: - host: grafana.example.org http: paths: - path: / pathType: Prefix backend: service: name: *name port: name: grafana - host: prometheus.example.org http: paths: - path: / pathType: Prefix backend: service: name: *name port: name: prometheus tls: - secretName: defaultgrafpro.sh//警告：请修改路径和访问域名 # 配置存储路径 export GRAFPRO_STORAGE=${GRAFPRO_STORAGE:-"/srv/grafpro"} # 配置访问域名 export GRAFANA_DOMAIN=${GRAFPRO_DOMAIN:-"grafana.example.org"} export PROMETHEUS_DOMAIN=${PROMETHEUS_DOMAIN:-"prometheus.example.org"} # 修改参数并部署服务 cat grafpro.yaml \ | sed "s#/srv/grafpro#$GRAFPRO_STORAGE#g" \ | sed "s#grafana.example.org#$GRAFANA_DOMAIN#g" \ | sed "s#prometheus.example.org#$PROMETHEUS_DOMAIN#g" \ | kubectl apply -f -部署chmod +x grafpro.sh ./grafpro.sh测试打开注意以下，开启端口9115、9090 浏览器打开地址http://grafana.example.org 账号密码都是admin，首次登录，提示修改密码，修改后自动跳到控制台浏览器打开http://grafana.example.org/connections/datasources/选择第一个，然后编辑URL为：http://127.0.0.1:9090 然后保存然后选择创建好的Prometheus，导入面板浏览器打开http://prometheus.example.org，查看信息配置Promethues任务//回到主节点的/srv/grafpro/etc目录下编辑yml文件，备份一下原有的yml，创建新的yml mv prometheus.yml prometheus00.yml //以下是yml文件内容（若部署时修改了负载名称blackbox-exporter，下文的配置文件也要做相应的修改） global: scrape_timeout: 15s scrape_interval: 1m evaluation_interval: 1m scrape_configs: # prometheus - job_name: prometheus static_configs: - targets: - 127.0.0.1:9090 # blackbox_all - job_name: blackbox_all static_configs: - targets: - blackbox-gz:9115 labels: region: '广州,腾讯云' # http_status_gz - job_name: http_status_gz metrics_path: /probe params: module: [http_2xx] #配置get请求检测 static_configs: - targets: - https://www.example.com labels: project: 测试1 desc: 测试网站描述1 - targets: - https://www.example.org labels: project: 测试2 desc: 测试网站描述2 basic_auth: username: ****** password: ****** relabel_configs: - target_label: region replacement: '广州,腾讯云' - source_labels: [__address__] target_label: __param_target - source_labels: [__param_target] target_label: instance - target_label: __address__ replacement: blackbox-gz:9115:80然后重启svc，方法如下：首先查看podkubectl get pod 然后删除查看到关于grafana的pod，然后稍等几分钟即可 kubectl delete pod *导入 Grafana 仪表盘下载附件json在Grafana仪表盘里导入即可导入后可以查看到监控仪已经开始了，显示各项信息

【k8s】将Go服务上传到k8s

2023-06-07T22:56:55+08:00

将Go服务发布到k8s集群首先安装go环境下载go的tar.gz包，可以前往阿里云镜像站go页面下载相应版本地址：https://mirrors.aliyun.com/golang/?spm=a2c6h.13651104.mirror-free-trial.1.75b41e57BOxyw5然后下载到虚拟机中，并且解压到/usr/local/src中//解压缩go包 [root@master ~]# tar -zxf go1.18.10.linux-amd64.tar.gz -C /usr/local/src //添加环境变量 [root@master src]# vim /etc/profile //添加如下： export GOROOT=/usr/local/src export PATH=$PATH:$GOROOT/bin //保存退出后source一下 source /etc/profile //查看是否成功 go version创建源码文件[root@master ~]# mkdir 0607tz [root@master ~]# cd 0607tz/ [root@master 0607tz]# vim main.go //编辑到文件中 package main import ( "net/http" "github.com/gin-gonic/gin" ) func statusOKHandler(c *gin.Context) { c.JSON(http.StatusOK, gin.H{"status": "success~welcome to study"}) } func versionHandler(c *gin.Context) { c.JSON(http.StatusOK, gin.H{"version": "v1.1版本"}) } func main() { router := gin.New() router.Use(gin.Recovery()) router.GET("/", statusOKHandler) router.GET("/version", versionHandler) router.Run(":8080") }Go mod初始化项目初始化项目[root@master 0607tz]# go mod init 0607tz go: creating new go.mod: module 0607tz go: to add module requirements and sums: go mod tidy //成功初始化 //设置代理 [root@master 0607tz]# go env -w GOPROXY=https://goproxy.cn,direct [root@master 0607tz]# go mod tidy //构建源码 [root@master 0607tz]# CGO_ENABLED=0 GOOS=linux GOARCH=amd64 go build -o k8s-demo main.go创建镜像编写dockerfile文件[root@master 0607tz]# vim Dockerfile FROM alpine ADD k8s-demo /data/app/ WORKDIR /data/app/ CMD ["/bin/sh","-c","./k8s-demo"]构建镜像[root@master 0607tz]# docker build -t taozheng/k8sdemo:v1 .打包镜像，传到k8s工作节点[root@master 0607tz]# docker save -o k8sdemo.tar.gz taozheng/k8sdemo:v1 [root@master 0607tz]# scp k8sdemo.tar.gz node:/root/ k8sdemo.tar.gz 100% 16MB 68.0MB/s 00:00 //在node节点解压镜像创建deployment的yaml文件[root@master 0607tz]# vim k8s.yaml //k8s.yaml apiVersion: apps/v1 kind: Deployment metadata: name: k8s-demo namespace: default labels: app: k8s-demo cy: taozheng spec: selector: matchLabels: app: k8s-demo replicas: 4 template: metadata: labels: app: k8s-demo spec: containers: - image: taozheng/k8sdemo:v1 imagePullPolicy: IfNotPresent name: k8s-demo ports: - containerPort: 8080 protocol: TCP resources: limits: cpu: 100m memory: 100Mi requests: cpu: 50m memory: 50Mi livenessProbe: tcpSocket: port: 8080 initialDelaySeconds: 10 timeoutSeconds: 3 readinessProbe: httpGet: path: / port: 8080 initialDelaySeconds: 10 timeoutSeconds: 2创建go的服务[root@master 0607tz]# vim gosvc.yaml kind: Service apiVersion: v1 metadata: name: k8s-demo-svc namespace: default labels: app: k8s-demo cy: taozheng spec: ports: - name: api port: 8080 protocol: TCP targetPort: 8080 selector: app: k8s-demok8s部署查看pod和服务信息修改svc的type类型变成nodePort[root@master 0607tz]# kubectl edit svc k8s-demo-svc保存后再次查看，已经修改成功查看svc标签浏览器测试访问：

【Hadoop】全分布式安装集群

2023-05-28T12:39:00+08:00

Hadoop全分布式安装环境准备首先做免密登录，三台虚拟机分别生成秘钥文件//三台都需要操作 ssh-keygen -t rsa //三台都需要打以下命令，进行秘钥分发 [root@tz1-123 ~]# ssh-copy-id tz1-123 [root@tz1-123 ~]# ssh-copy-id tz2-123 [root@tz1-123 ~]# ssh-copy-id tz3-123修改防火墙设置后，查看三台虚拟机防火墙状态//三台，禁止防火墙自启，立即生效 systemctl disable firewalld --now安装JDK(3台)，首先上传安装包到根目录，然后解压在/usr/lib/jvm//修改环境变量/etc/profile JAVA_HOME=/usr/lib/jvm/jdk1.8.0_152 JRE_HOME=$JAVA_HOME/jre CLASS_PATH=.:$JAVA_HOME/lib/dt.jar:$JAVA_HOME/lib/tools.jar:$JRE_HOME/lib PATH=$PATH:$JAVA_HOME/bin:$JRE_HOME/bin export JAVA_HOME JRE_HOME CLASS_PATH PATH部署首先将Hadoop软件包上传至/root中，并解压在/usr/local/src/下修改core-site.xml//将以下内容加在之间 fs.defaultFS hdfs://tz1-123:9000 hadoop.tmp.dir /usr/local/src/hadoop/data/tmp 修改hadoop-env.sh//在最后添加以下一条 export JAVA_HOME=/usr/lib/jvm/jdk1.8.0_152修改hdfs-site.xml//将以下内容加在之间 dfs.replication 3 dfs.namenode.secondary.http-address tz3-123:50090 修改yarn-env.sh//在最后添加以下一条 export JAVA_HOME=/usr/lib/jvm/jdk1.8.0_152修改yarn-site.xml//将以下内容加在之间 yarn.nodemanager.aux-services mapreduce_shuffle yarn.resourcemanager.hostname tz2-123 修改mapred-env.sh//在最后添加以下一条 export JAVA_HOME=/usr/lib/jvm/jdk1.8.0_152修改mapred-site.xml[root@tz1-123 hadoop]# cp mapred-site.xml.template mapred-site.xml [root@tz1-123 hadoop]# vim mapred-site.xml //将以下内容加在之间 mapreduce.framework.name yarn 修改slaves[root@tz1-123 hadoop]# vim slaves tz1-123 tz2-123 tz3-123分发Hadoop软件包[root@tz1-123 hadoop]# scp -r /usr/local/src/hadoop tz2-123:/usr/local/src/ [root@tz1-123 hadoop]# scp -r /usr/local/src/hadoop tz3-123:/usr/local/src/修改/etc/profile//在最后添加以下内容 export HADOOP_HOME=/usr/local/src/hadoop export PATH=$PATH:$HADOOP_HOME/bin:$HADOOP_HOME/sbin //修改完成后再次分发，并加在环境变量 [root@tz1-123 hadoop]# source /etc/profile [root@tz1-123 hadoop]# scp /etc/profile tz2-123:/etc/ [root@tz1-123 hadoop]# scp /etc/profile tz3-123:/etc/在tz1-123（master）上格式化namenodehdfs namenode -format启动集群并测试[hadoop@tz1-123 ~]$ start-dfs.sh [hadoop@tz2-123 ~]$ start-yarn.sh [root@tz1-123 hadoop]# jps 8096 NameNode 24690 NodeManager 24882 Jps 8293 DataNode [root@tz2-123 ~]# jps 30709 NodeManager 24086 DataNode 30567 ResourceManager 781 Jps [root@tz3-123 ~]# jps 23988 DataNode 604 Jps 30494 NodeManagerHDFS Shell操作HDFS Shell操作以hadoop fs或hdfs dfs开头。#查看某路径下文件夹 hadoop fs -ls HDFS路径 #在HDFS上创建文件夹 hadoop fs -mkdir HDFS文件夹路径 #在HDFS上创建文件夹（上级目录不存在） hadoop fs -mkdir -p HDFS文件夹路径 #将本地文件上传到HDFS上 hadoop fs -put 本地文件路径 HDFS路径 #查看集群文件的内容 hadoop fs -cat HDFS文件路径 #从HDFS上下载文件到本地 hadoop fs -get HDFS文件路径本地路径 #删除HDFS上空文件夹 hadoop fs -rmdir HDFS文件夹路径 #删除HDFS上的非空文件夹 hadoop fs -rm -r HDFS文件夹路径 #删除HDFS上的文件 hadoop fs -rm HDFS文件路径 #将HDFS上的路径剪切至另一个路径下 hadoop fs -mv HDFS源路径 HDFS目标路径 #将HDFS上的路径复制到另一个路径下 hadoop fs -cp HDFS源路径 HDFS目标路径 #在HDFS上创建一个文件 hadoop fs -touchz HDFS路径我的博客即将同步至腾讯云开发者社区，邀请大家一同入驻：https://cloud.tencent.com/developer/support-plan?invite_code=35n3trqr2ug48

【Docker】k8s健康检查

2023-05-28T12:05:24+08:00

健康检查使用存活探针创建使用 execaction 模式的存活探针 pod 的 yaml 文件。需要创建目录（/tmp/healthy）查看到运行成功，持续监控pod状态，看到pod反复重启使用 describe 命令查看详细 pod 信息，正常创建使用 http 存活探针的 pod 的 yaml 文件。创建yaml运行并查看状态查看详细events创建使用 tcp 存活探针的 pod 的 yaml，模板采用 httpd 容器镜像。创建yaml文件运行并进行容器内操作查看pod的restarts次数查看pod之前未通过liveness的记录就绪探针创建 http 的 deployment 的 yaml 文件，其中配置 readiness 探针。运行deployment使用describechakanhttp服务的endpoint可以看到有4个地址进入一个容器，删除index.html文件再使用 describe 命令查看 endpoint可以看到删除的pod地址已经从endpoint中移除查看pod的详细信息，看到pod未通过探针检测查看pod信息，kandaopod处于notready状态

【k8s】标签Label与Label Selector

2023-05-16T20:43:42+08:00

Label与Label Selector标签进入目录保存实验文件并创建一个yaml使用多个标签[root@master tz123]# cd /root/tz123/labfile/labelfile [root@master labelfile]# vim labelpod.yaml kind: Pod apiVersion: v1 metadata: name: labelpod labels: app: busybox version: new spec: containers: - name: labelpod image: busybox args: - /bin/sh - -c - sleep 30000创建Pod，并查看pod的label[root@master labelfile]# kubectl apply -f labelpod.yaml pod/labelpod created [root@master labelfile]# kubectl get pod --show-labels NAME READY STATUS RESTARTS AGE LABELS labelpod 1/1 Running 0 11s app=busybox,version=new为容器添加新标签[root@master labelfile]# kubectl label pod labelpod time=2019 pod/labelpod labeled [root@master labelfile]# kubectl get pod --show-labels NAME READY STATUS RESTARTS AGE LABELS labelpod 1/1 Running 0 69s app=busybox,time=2019,version=new标签选择器创建新的yaml[root@master labelfile]# vim labelpod2.yaml kind: Pod apiVersion: v1 metadata: name: labelpod2 labels: app: httpd version: new spec: containers: - name: httpd image: httpd创建并查看新创建的labelpod2[root@master labelfile]# kubectl apply -f labelpod2.yaml [root@master labelfile]# kubectl get pod --show-labels NAME READY STATUS RESTARTS AGE LABELS labelpod 1/1 Running 0 12m app=busybox,time=2019,version=new labelpod2 0/1 ContainerCreating 0 23s app=httpd,version=new使用给予等值的标签选择器[root@master labelfile]# kubectl get pod -l app=httpd NAME READY STATUS RESTARTS AGE labelpod2 1/1 Running 0 100s 或 [root@master labelfile]# kubectl get pod -l app==httpd NAME READY STATUS RESTARTS AGE labelpod2 1/1 Running 0 114s使用基于不等值的标签选择器和查看pod针对某标签键的值[root@master labelfile]# kubectl get pod -l app!=httpd NAME READY STATUS RESTARTS AGE labelpod 1/1 Running 0 14m [root@master labelfile]# kubectl get pod -L app NAME READY STATUS RESTARTS AGE APP labelpod 1/1 Running 0 15m busybox labelpod2 1/1 Running 0 3m5s httpd使用标签选择器实现调度将节点1打上标签并查看[root@master labelfile]# kubectl label node node env=test node/node labeled [root@master labelfile]# kubectl get node -L env NAME STATUS ROLES AGE VERSION ENV master Ready control-plane,master 91d v1.20.6 node Ready 91d v1.20.6 test使用nodeselector实现调度，创建新的yaml文件[root@master labelfile]# vim nsdeploy.yaml kind: Deployment apiVersion: apps/v1 metadata: name: nginx-dy labels: app: nginx spec: replicas: 3 selector: matchLabels: app: nginx template: metadata: labels: app: nginx spec: containers: - name: nginx image: nginx:1.7.9 ports: - containerPort: 80 nodeSelector: env: test查看deployment中的pod位置[root@master labelfile]# kubectl get pod -o wide NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES labelpod 1/1 Running 0 28m 10.244.167.145 node labelpod2 1/1 Running 0 15m 10.244.167.146 node nginx-dy-6dd6c76bcb-667ss 1/1 Running 0 5m19s 10.244.167.148 node nginx-dy-6dd6c76bcb-q8tqh 1/1 Running 0 5m19s 10.244.167.149 node nginx-dy-6dd6c76bcb-xc9h7 1/1 Running 0 5m19s 10.244.167.147 node 使用 node affinity 调度，创建一个新的 yaml 文件 nadeploy2.yaml[root@master labelfile]# vim nadeploy2.yaml kind: Deployment apiVersion: apps/v1 metadata: name: httpd-dy labels: app: httpd spec: replicas: 3 selector: matchLabels: app: httpd template: metadata: labels: app: httpd spec: containers: - name: httpd image: httpd affinity: nodeAffinity: requiredDuringSchedulingIgnoredDuringExecution: nodeSelectorTerms: - matchExpressions: - key: env operator: In values: - test创建deployment并查看deployment中的pod位置，三个pod都在node上[root@master labelfile]# kubectl apply -f nadeploy2.yaml deployment.apps/httpd-dy created [root@master labelfile]# kubectl get pod -o wide NAME READY STATUS RESTARTS AGE IP NODE NOMINATED NODE READINESS GATES httpd-dy-5b4bb9646-g4jzb 1/1 Running 0 33s 10.244.167.150 node httpd-dy-5b4bb9646-lb876 1/1 Running 0 33s 10.244.167.151 node httpd-dy-5b4bb9646-q7zcm 0/1 ContainerCreating 0 33s node labelpod 1/1 Running 0 38m 10.244.167.145 node labelpod2 1/1 Running 0 26m 10.244.167.146 node nginx-dy-6dd6c76bcb-667ss 1/1 Running 0 15m 10.244.167.148 node nginx-dy-6dd6c76bcb-q8tqh 1/1 Running 0 15m 10.244.167.149 node nginx-dy-6dd6c76bcb-xc9h7 1/1 Running 0 15m 10.244.167.147 node 实训任务创建一个deployment使用镜像nginx，5个副本deployment中的pod不能出现在node上[root@master labelfile]# vim shixun01.yaml kind: Deployment apiVersion: apps/v1 metadata: name: nginx-dy labels: app: nginx spec: replicas: 1 selector: matchLabels: app: nginx template: metadata: labels: app: nginx spec: containers: - name: nginx image: nginx affinity: nodeAffinity: requiredDuringSchedulingIgnoredDuringExecution: nodeSelectorTerms: - matchExpressions: - key: env operator: Not In values: - node寻找一种方式搜索出kubernetes系统上提供core-dns，kubeproxy以及dashboard服务的pod先给core-dns，kubeproxy，dashboard打上标签[root@master labelfile]# kubectl label -n kube-system pod kube-proxy-kj8j5 app=kubeproxy [root@master labelfile]# kubectl label -n kube-system pod coredns-7f89b7bc75-n224r app=coredns查找关键词的pod搜索dashboard的pod使用标签和标签选择器，使用一条命令删除node2节点的nginx

【Hive】Hadoop下的部署（未上接）

2023-03-27T23:41:04+08:00

Hive的部署MySQL安装安装首先上传mysql数据库的rpm压缩包到主机/opt/software//解压缩包到当前目录中 [root@master-tz software]# unzip mysql-5.7.18.zip //进入到rpm软件目录中，首先检查mariadb，如果有就要卸载//安装数据库 [root@master-tz mysql-5.7.18]# rpm -ivh mysql-community-common-5.7.18-1.el7.x86_64.rpm [root@master-tz mysql-5.7.18]# rpm -ivh mysql-community-libs-5.7.18-1.el7.x86_64.rpm [root@master-tz mysql-5.7.18]# rpm -ivh mysql-community-client-5.7.18-1.el7.x86_64.rpm //安装下面的rpm需要首先安装perl软件 [root@master-tz mysql-5.7.18]# yum install -y net-tools perl [root@master-tz mysql-5.7.18]# rpm -ivh mysql-community-server-5.7.18-1.el7.x86_64.rpm配置修改配置文件/etc/my.cnfvim /etc/my.cnf //最后添加五行 default-storage-engine=innodb innodb_file_per_table collation-server=utf8_general_ci init-connect='SET NAMES utf8' character-set-server=utf8 //最后保存退出启动[root@master-tz mysql-5.7.18]# systemctl start mysqld [root@master-tz mysql-5.7.18]# systemctl status mysqld查看MySQL初始密码[root@master-tz mysql-5.7.18]# cat /var/log/mysqld.log | grep password 2023-03-27T08:52:43.074230Z 1 [Note] A temporary password is generated for root@localhost: KbVXiHlul3:> //查看初始密码，下方需要填写 [root@master-tz mysql-5.7.18]# mysql_secure_installation //重新设定密码，并把密码设置为Password123$ //注：允许远程连接设定为n，表示允许远程连接，其它设定为y除了以下是n其他都是y登录数据库客户端[root@master-tz mysql-5.7.18]# mysql -uroot -pPassword123$新建hive用户与元数据mysql>create database hive_db; mysql>create user hive identified by 'Password123$'; mysql>grant all privileges on *.* to hive@'%' identified by 'Password123$' with grant option ; mysql>grant all privileges on *.* to 'root'@'%'identified by 'Password123$' with grant option; mysql>flush privileges;Hive安装安装首先将hive的压缩包上传到虚拟机，并解压，重命名hive，设置hive权限[root@master-tz ~]# tar -zxf apache-hive-2.0.0-bin.tar.gz -C /usr/local/src/ [root@master-tz ~]# cd /usr/local/src/ [root@master-tz src]# mv apache-hive-2.0.0-bin/ hive [root@master-tz src]# chown -R hadoop:hadoop hive/修改环境变量[root@master-tz src]# vim /etc/profile # set Hive environment export HIVE_HOME=/usr/local/src/hive # Hive安装目录 export PATH=$HIVE_HOME/bin:$PATH # 添加将Hive的bin目录 export HIVE_CONF_DIR=$HIVE_HOME/conf #Hive的环境变量 [root@master-tz src]# source /etc/profile修改配置文件hive-site.xml文件首先切换到hadoop用户[hadoop@master-tz conf]$ cd /usr/local/src/hive/conf [hadoop@master-tz conf]$ vim hive-site.xml javax.jdo.option.ConnectionURL jdbc:mysql://master-tz:3306/hive_db?createDatabaseIfNotExist=true javax.jdo.option.ConnectionUserName hive javax.jdo.option.ConnectionPassword Password123$ javax.jdo.option.ConnectionDriverName com.mysql.jdbc.Driver hive.downloaded.resources.dir /usr/local/src/hive/tmp hive.exec.local.scratchdir /usr/local/src/hive/tmp/${hive.session.id}_resources hive.querylog.location /usr/local/src/hive/tmp hive.server2.logging.operation.log.location /usr/local/src/hive/tmp/operation_logs hive.server2.webui.host master-tz hive.server2.webui.port 10002 hive-env.sh[hadoop@master-tz conf]$ cp hive-env.sh.template hive-env.sh [hadoop@master-tz conf]$ vim hive-env.sh //增加如下配置项 # Set JAVA export JAVA_HOME=/usr/local/src/java # Set HADOOP_HOME to point to a specific hadoop install directory export HADOOP_HOME=/usr/local/src/hadoop # Hive Configuration Directory can be controlled by: export HIVE_CONF_DIR=/usr/local/src/hive/conf # Folder containing extra ibraries required for hive compilation/execution can be controlled by: export HIVE_AUX_JARS_PATH=/usr/local/src/hive/lib将MySQL的驱动jar包上传至虚拟机，然后将该jar包复制到hive安装路径下的lib文件夹中[root@master-tz software]# cp mysql-connector-java-5.1.46.jar /usr/local/src/hive/lib/确保hadoop集群正常，然后初始化hive元数据库[hadoop@master-tz conf]$ schematool -initSchema -dbType mysql进入hive shell界面[hadoop@master-tz ~]$ hive hive>如果出现以下情况则需要去hive-site.xml配置文件修改为jdbc:mysql://master-tz:3306/hive_db?createDatabaseIfNotExist=true&useSSL=false

kubeadm方式部署k8s集群

2022-12-07T08:58:00+08:00

环境准备第一台节点（主节点）： 192.168.123.200 master第二台节点（从节点）： 192.168.123.201 slave以下文件需要单独下载云盘地址地址配置/etc/hosts域名解析（两台）[root@master ~]# cat /etc/hosts 127.0.0.1 localhost localhost.localdomain localhost4 localhost4.localdomain4 ::1 localhost localhost.localdomain localhost6 localhost6.localdomain6 192.168.123.200 master-tz 192.168.123.201 slave01-tz关闭防火墙、SELINUX和swap（两台）1.systemctl disable firewalld --now 2.setenforce 0 3./etc/selinux/config中的一行修改为SELINUX=disabled 4.swapoff -a 5./etc/fstab中的swap加注释 #/dev/mapper/centos-swap swap swap defaults 0 0配置系统内核参数，使流过网桥的流量页进入IPTables/Netfilter（两台）[root@master ~]# modprobe br_netfilter [root@master ~]# echo "modprobe br_netfilter" >> /etc/profi le [root@master ~]# cat > /etc/sysctl.d/k8s.conf < --discovery-token-ca-cert-hash sha256:d6a0bb61368c23be10444d7a18eab071b750c97c45186020980714fd57b13bdd再次查看master节点[root@master-tz ~]# kubectl get nodes NAME STATUS ROLES AGE VERSION master-tz NotReady control-plane,master 8m31s v1.20.6 slave01-tz NotReady 12s v1.20.6此时集群状态还是NotReady 状态，因为没有安装网络插件。若要扩充集群（master节点）master执行 kubeadm token create --print-join-command 结果在新增节点执行安装k8s网络组件Calico（master节点）上传calico.yaml文件到master节点。[root@master ~]# kubectl apply -f calico.yaml 再次使用kubectl get nodes命令查看节点状态为Ready安装dashboard（master节点）上传dashboard_2_0_0.tar.gz和metrics-scrapter-1-0-1.tar.gz到两个节点。上传kubernetes-dashboard.yaml到master节点。[root@master ~]# docker load -i dashboard_2_0_0.tar.gz [root@master ~]# docker load -i metrics-scrapter-1-0-1.tar.gz [root@master ~]# kubectl apply -f kubernetes-dashboard.yaml [root@master-tz ~]# kubectl get pods -n kubernetes-dashboard NAME READY STATUS RESTARTS AGE dashboard-metrics-scraper-7445d59dfd-p572g 1/1 Running 0 10s kubernetes-dashboard-54f5b6dc4b-5zxpm 1/1 Running 0 10s 说明dasbnoard安装成功了。查看dashboard的service[root@master-tz ~]# kubectl get svc -n kubernetes-dashboard NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE dashboard-metrics-scraper ClusterIP 10.101.56.238 8000/TCP 97s kubernetes-dashboard ClusterIP 10.97.126.230 443/TCP 97s修改service type的类型为NodePort[root@master ~]# kubectl edit svc kubernetes-dashboard -n kubernetes-dashboard[root@master ~]# kubectl get svc -n kubernetes-dashboard通过浏览器进行访问https://192.168.123.200:30245通过token登录dashboard（master节点）创建管理员token，具有查看任何空间的权限，可以管理所有资源对象。[root@master-tz ~]# kubectl create clusterrolebinding dashboard-cluster-admin --clusterrole=cluster-admin --serviceaccount=kubernetes-dashboard:kubernetes-dashboard clusterrolebinding.rbac.authorization.k8s.io/dashboard-cluster-admin created [root@master-tz ~]# kubectl get secret -n kubernetes-dashboard NAME TYPE DATA AGE default-token-scvqs kubernetes.io/service-account-token 3 14m kubernetes-dashboard-certs Opaque 0 14m kubernetes-dashboard-csrf Opaque 1 14m kubernetes-dashboard-key-holder Opaque 2 14m kubernetes-dashboard-token-bs98s kubernetes.io/service-account-token 3 14m [root@master-tz ~]# kubectl describe secret kubernetes-dashboard-token-bs98s -n kubernetes-dashboard Name: kubernetes-dashboard-token-bs98s Namespace: kubernetes-dashboard Labels: Annotations: kubernetes.io/service-account.name: kubernetes-dashboard kubernetes.io/service-account.uid: d0842b14-e79e-4129-b6e3-bfd3c7039334 Type: kubernetes.io/service-account-token Data ==== ca.crt: 1066 bytes namespace: 20 bytes token: eyJhbGciOiJSUzI1NiIsImtpZCI6IkdHTy1lQ2tndl9qQ29INUtEMEREMW1iUWhWeENOODB1Q2lOOERSYnN6OTQifQ.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.f6iGY-QbB5YQFuaTkU6qR9UBTbFiIcDbpgT40E_ceQZGh3kdWyKzeTB-pWUkrJV1gWFaQt3Er7_brB-T7juO8eywunXkE6Xd_xH7XzaiWbNYFYfr3gMMXI8SmbnpqDKHclqw_tUIgun37ao7YYY_22_mYDdcTSIVFvx9XehK48eJWVfdyy-snuZiTKoR2pKMH0Rau3oXKlw7is8bV7yezeucZnaMPa60N-1KIMAvRM7gXlMX9m_BKiqvxEoru-2FDEoOkiCFXV-juGclxM_Qtn70i9R2JVjPgE5VX_gP7RFHDoXIEwykyjJqOg2fguE9Vy8nKnrfOo0c99aGXxnW_g使用token值登录测试创建nginx拉取nginx镜像docker pull nginx创建nginx应用服务kubectl create deployment ngix-deployment1 --image nginx --port=80 --replicas=2创建service服务kubectl expose deployment ngix-deployment1 --name=nginx --port=80 --target-port=80 --type=NodePort访问nginx服务[root@master-tz ~]# kubectl get svc NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE kubernetes ClusterIP 10.96.0.1 443/TCP 109m nginx NodePort 10.110.242.218 80:31079/TCP 9s

【swarm】docker环境下的集群服务

2022-11-09T11:38:00+08:00

介绍Docker Swarm是Docker原生的集群工具，因而无须使用额外的编排软件创建或管理集群。DockerSwarm部署更简单，适合规模不大的应用程序环境，尤其适用于简单和快速开发。Docker Client：客户端 Swarm Manager：管理器节点 Scheduler：调度器Discovery Service：服务发现 Swarm Node：工作者节点 Docker Containers：容器基本命令集群管理命令docker swarm ca：显示和轮转根CA。 docker swarm init：初始化集群。 docker swarm join：作为节点加入集群。 docker swarm join-token：管理加入集群的令牌。 docker swarm leave：脱离集群。 docker swarm unlock：解锁集群。 docker swarm unlock-key：管理解锁密钥。 docker swarm update：更新集群节点管理命令docker node demote：将一个或多个管理器节点降级为工作者节点。 docker node inspect：显示一个或多个节点的详细信息。 docker node ls：列出Swarm集群中的节点。 docker node promote：将一个或多个节点升级为管理器节点。 docker node ps：列出在一个或多个节点（默认为当前节点）上运行的任务。 docker node rm：从Swarm集群中删除一个或多个节点。 docker node update：更新节点的选项，如可用性、标签或角色。实践操作主机名ip角色Manager192.168.123.100主控Worker01192.168.123.101节点01Worker02192.168.123.102节点02初始化swarm集群在管理器节点上执行以下命令获取加入管理器角色节点的命令（含令牌）docker swarm init --advertise-addr 192.168.123.100添加节点将上操作中的代码复制到worker01、02主机上查看集群信息docker node list

【ansible】linux下的集群管理服务

2022-11-08T17:33:00+08:00

介绍ansible是新出现的自动化运维工具，基于Python开发，集合了众多运维工具（puppet、cfengine、chef、func、fabric）的优点，实现了批量系统配置、批量程序部署、批量运行命令等功能。Ansible架构相对比较简单，仅需通过SSH连接客户机执行任务即可主机名ip角色Server192.168.123.195主控Backend01192.168.123.196被控01Backend02192.168.123.197被控02安装ansible准备eple源这边使用的是阿里云的源wget -O /etc/yum.repos.d/epel.repo https://mirrors.aliyun.com/repo/epel-7.repo下载完成后就可以安装ansible服务yum install -y ansible openssh-clients配置ansible修改配置文件 vim /etc/ansible/ansible.cfg //* 71 行，取消注释。开启关闭对未连接的主机 SSH 秘钥检测 host_key_checking = False编写 Ansible 主机管理文件mv /etc/ansible/hosts /etc/ansible/hosts.bak ##先备份一下内容 vi /etc/ansible/hosts ##建立目标群组，并写入被管机(目标服务器)的 IP/FQDN [tz1101] backend01 backend02测试连通性ansible tz1101 -m pingTip：配置ssh免密才可以出现上面结果的配置ssh免密登陆ssh-keygen一路回车，默认免密通信ssh-copy-id ip/主机名把密钥发送到集群主机中另外在需要/etc/hosts中写入主机名和ipansible应用添加用户为集群主机添加单独用户ansible tz1101 -m user -a 'name=tao state=present'添加完成用户后再设置一个密码，在server端用pip python 生成哈希密码首先安装pip pythonyum install python-pip -y生成密码ansible tz1101 -m user -a 'name=tao password=tarRU/F9EJjRU update_password=always'查看一下playbook剧本测试集群安装一个httpd软件的playbook剧本vim playbook_create_install.yml #编写剧本文件→安装软件 - hosts: tz1101 #集群组名 tasks: - name: install vsftpd ##任务名称 yum: name=vsftpd state=installed ##安装vsftpd - name: running and enabled service: name=vsftpd state=started enabled=yes ##设置开机自启动使用tag标签创建tag文件yml注：自定义了tag，他就会只执行带有tag=test2的内容其他标签内容不会执行使用变量自定义变量安装服务- hosts: tz1101 become: yes become_method: sudo tasks: - name: installed bao yum: name={{ item }} state=installed with_items: - vim-enhanced ##软件名 - wget - unzip tags: Taozhengansible-playbook bianliang.ymlroles内网中需要关闭防火墙和selinuxansible tz1101 -a "setenforce 0 && systemctl stop firewalld"在当前server主机内安装tree服务yum install -y tree群组下的主机也要安装tree软件ansible tz1101 -m yum -a "name=tree state=installed"/*-m 使用模块 yum命令 -a是具体内容 name是tree state是操作安装创建roles子目录及内容mkdir -p roles/ins_httpd/{files,tasks,vars}回到roles，编写一个yml文件Vim playbook_httpd.yml - hosts: tz1101 roles: - ins_httpd 编写vars下的main文件 vim roles/ins_httpd/vars/main.yml packages: - httpd创建任务剧本编写tasks下的main文件 vim roles/ins_httpd/tasks/main.yml - name: httpd is installed yum: name=httpd state=installed tags: install_httpd - name: edit httpd.conf lineinfile: > dest=/etc/httpd/conf/httpd.conf regexp="{{item.regexp}}" line="{{item.line}}" with_items: - { regexp: "^#ServerName",line: "ServerName {{ansible_fqdn}}:80" } tags: edit_httpd.conf - name: httpd is running and enabled service: name=httpd state=started enabled=yes - name: put index.html copy: src=index.html dest=/var/www/html owner=root group=root mode=0644执行剧本ansible-playbook playbook_httpd.yml查看写入的httpd的index.htmlansible tz1101 -m shell -a "curl localhost"

【Packmaker】集群部署实践+测试

2022-10-19T08:13:00+08:00

主机名ip角色Node01192.168.123.23节点1Node02192.168.123.123节点2首先安装pssh远程服务yum install -y pssh两台虚拟机都需要编辑hosts文件分别写入node01、node02的ip 和主机名生成免密登录，就可以无需密码直接登陆node01和node02ssh-keygen -t rsassh-copy-id -i /root/.ssh/id_rsa.pub root@node02-tzNode02端也需要进行相同操作之后安装pacemaker服务并设置开机自启pssh -h host-list.txt -i 'yum install pacemaker pcs -y'pssh -h host-list.txt -i 'systemctl enable --now pcsd'配置 Pacemaker Cluster：设置cluster的账号密码pssh -h host-list.txt -i 'echo 123456 | passwd --stdin hacluster'设置防火墙：pssh -h host-list.txt -i 'firewall-cmd --add-service=high-availability --permanent'添加服务pssh -h host-list.txt -i 'firewall-cmd --reload'加载防火墙建立 Cluster 授权：pcs cluster auth node01-tz node02-tz配置 Cluster(集群名称为:tz-cluster)：pcs cluster setup --name tz-cluster node01-tz node02-tz启动 Cluster:pcs cluster start –all使 Cluster 随机启动:pcs cluster enable --all验证集群状态:pcs status cluster添加集群资源:两台节点需要安装httpd服务，不需要启动pssh -h host-list.txt -i 'yum install httpd -y'两台服务器都需要编辑一个新的配置文件vim /etc/httpd/conf.d/server_status.confExtendedStatus On SetHandler server-status Require local 并写入一个测试页面到/var/www/html/index.html配置 Cluster 并设定 VIP：[root@node01-tz ~]# pcs property set stonith-enabled=false <==关闭 stonith 功能[root@node01-tz ~]# pcs property set no-quorum policy=ignore <==忽略仲裁[root@node01-tz ~]# pcs property set default-resource-stickiness="INFINITY" <==设置资源超时时间[root@node01-tz ~]# pcs property list<==显示设定pcs resource create VIP ocf:heartbeat:IPaddr2 ip=192.168.123.111 cidr_netmask=24 op monitor interval=30s <==设定 VIP pcs status resources <==显示资源信息添加 httpd 资源: pcs resource create Web123-Cluster ocf:heartbeat:apache configfile=/etc/httpd/conf/httpd.conf statusurl="http://127.0.0.1/server-status" op monitor interval=1min[root@node01-tz ~]# pcs constraint colocation add Web123-Cluster with VIP INFINITY[root@node01-tz ~]# pcs constraint order VIP then Web123-Cluster <==设定启动顺讯:先启动 VIP 在启动 WEB123-Clusterpcs constraint <==显示资源情况客户端测试:停止 ndoe1 节点的 httpd 资源,客户端再测试:pcs cluster stop node01-tz就自动切换到了node02上使用 WEB GUI 管理 PaceMakerfirewall-cmd --add-port=2224/tcp --permanentfirewall-cmd --reload访问https://192.168.123.111:2224注：①80端口占用查询：netstat -tlnp|grep 80杀死已经存在的80端口服务Kill -9 进程号firewall-cmd --zone=public--add-port=80/tcp --permanent ###添加80端口Firewall-cmd --reload ###重载防火墙②ssh密钥生成信息只能存在一个，需要手动添加到本机与对方的auth开头文件和konw文件中>>>>>/root/.ssh/