u010533742的博客

环境说明：1、kubekey 2.2.12、kubesphere 3.3.03、kubernetes 1.23.74、kube-prometheus-stack 35.0.0官方文档https://kubesphere.com.cn/docs/v3.3/faq/observability/byop/要使用您自己的 Prometheus 堆栈设置，请执行以下步骤：步骤1、卸载 KubeSphere 的自定义 Prometheus 堆栈2、安装您自己的 Prometheus 堆栈KubeSph

1.2 生成manifest-sample.yaml#修改manifest-sample.yaml1.2导出并生成离线安装包2.2 生成离线集群配置文件#修改config-sample.yaml文件3.1、创建私仓库3.2创建仓库的projectbash create_project_harbor.sh4、 验证https://blog.youkuaiyun.com/ma726518972/article/details/125077760.........

https://zhuanlan.zhihu.com/p/415594861https://zhuanlan.zhihu.com/p/379361295https://www.cnblogs.com/larrydpk/p/14953946.html

一、如果使用metallb的layer2模式，需要如kube-proxy进行配置、让metallb接管主机上的arp管理。1、POD部暑kube-proxykubectl edit configmap -n kube-system kube-proxyconfigmap/kube-proxy editedapiVersion: kubeproxy.config.k8s.io/v1alpha1kind: KubeProxyConfigurationmode: "ipvs"ipvs: s

参考： https://blog.youkuaiyun.com/easylife206/article/details/124507103一、 架构图二、 CRD 简介Fluent Operator 为 Fluent Bit 和 Fluentd 分别定义了两个 Group：fluentbit.fluent.io 和 fluentd.fluent.io。fluentbit.fluent.io 分组下包含以下 6 个 CRDs：Fluentbit CRD 定义了 Fluent Bit 的属性，比如镜像版本、

https://doc.traefik.io/traefik/middlewares/目标：1、traefik与docker集成，docker通过使用label通知traefik进行配置2、集成https-redirect、basic-auth、digestauth等中间件3、docker network create traefikmkdir /data/traefik -pcd /data/traefikcat > docker-compose.yaml << 'E

一、promeprometheus: extraSecret: #配置thanos的bucket config, 里面包括了minio的配置 name: bucket-config data: bucket_config.yaml: | type: S3 config: bucket: "thanos-buckets01" #minio的桶名 end

traefik详细说明：https://www.qikqiak.com/traefik-book/一、添加traefik仓库#添加 Traefik v2 helm charthelm repo add traefik https://helm.traefik.io/traefik# 更新下仓库helm repo update #查询repohelm repo listtraefik https://helm.traefik.io/traefik二、部暑traefik2

一、安装docker run -d \--name admin-mongo \--restart always \-p 8001:1234 \-e HOST=0.0.0.0 \-e PORT=1234 \-e PASSWORD=123456 \-v /etc/localtime:/etc/localtime \mrvautin/adminmongo二、admin-mongo默认开启了login-auth# default login config"username":"admin"

一、在arm64下运行异常现象<--- Last few GCs ---><--- JS stacktrace --->## Fatal process OOM in insufficient memory to create an Isolate#二、排查方法进行容器中查看，直接运行node时同样报错(node版本为12)，说明node不行正常运行。docker run -d -it --rm --entrypoint /bin/sh --name mongo-

cat /run/flannel/subnet.env FLANNEL_NETWORK=172.120.0.0/16FLANNEL_SUBNET=172.120.1.1/24FLANNEL_MTU=1450FLANNEL_IPMASQ=trueoverlay 同一网段下vxlan、udp二、Pod之间互相访问1.flanneld将本机获取的subnet以及主机之间通信的public ip通过etcd存储起来,需要时发给相应模块2.flanneld通过各种backend mechani

1、后台运行容器的piddocker inspect etcd | grep -i pid2、查看cgroupmount -t cgroup#1、memorycd /sys/fs/cgroup/memory/3、docker权限docker容器权限设置–cap-add | --cap-drop | --privilegedCAP_CHOWN:修改文件属主的权限CAP_DAC_OVERRIDE:忽略文件的DAC访问限制CAP_DAC_READ_SEARCH:忽略文件读及目录搜索

#创建manifest,以gitlab-ce为例，包括两个镜像yrzr/gitlab-ce-arm64v8:14.4.2-ce.0(arm64)和gitlab/gitlab-ce:14.4.2-ce.0（amd64）docker manifest create zengxiangbang/gitlab-ce:14.4.2-ce.0 yrzr/gitlab-ce-arm64v8:14.4.2-ce.0 gitlab/gitlab-ce:14.4.2-ce.0#上传manifestdocker mani

mkdir -p /datagit clone https://github.com/goharbor/harbor.gitcd harborgit checkout -b v2.2.2 v2.2.2cd makecat > download_images.sh << 'EOF'namespace=cocl666version=v2.2.2imgs=(redis-photonharbor-dbharbor-jobserviceharbor-lognginx-p

git clone https://github.com/nacos-group/nacos-docker.gitdocker-compose -f example/standalone-mysql-5.7.yaml up -d访问url:http://xxx.xxx.xxx.xxx:8848/nacos/帐号，密码：nacos/nacos

一、添加cert-manager 仓库#添加 kubernetes-dashboard helm charthelm repo add jetstack https://charts.jetstack.io# 更新下仓库helm repo update #查询repohelm repo list 二、部暑cert-manager#指定变量pro=cert-managerchart_version=v1.5.3mkdir -p /data/$procd /data/$pro

一、添加traefik仓库#添加 Traefik v2 helm charthelm repo add traefik https://containous.github.io/traefik-helm-chart# 更新下仓库helm repo update #查询repohelm repo listtraefik https://containous.github.io/traefik-helm-chart 二、部暑traefik2#mkdir -p /data

mkdir /opt/bin -p#下载helm3cd /opt/binwget https://get.helm.sh/helm-v3.6.3-linux-amd64.tar.gz#把helm3解压到本目录下tar zxvf helm-v3.6.3-linux-amd64.tar.gz --strip-components 1 linux-amd64/helm#通过软链把helm链接到目录/usr/bin下ln -s /opt/bin/helm /usr/bin/helm..

进到etcd容器,以删除rook-ceph为例etcdctl del /registry/namespaces/rook-ceph也可以通这招删除podetcdctl del /registry/pods/default/nginx-test-0

#创建dockerfilecat > Dockerfle << 'EOF'FROM openjdk:11.0.12-jre-busterADD target/*.jar /EXPOSE 80 ENTRYPOINT ["/bin/sh","-c","java $JAVA_OPTS -XX:NativeMemoryTracking=detail -Djava.security.egd=file:/dev/./urandom -jar -Dserver.port=80 -Dloggin

yum install wget tar -y# 下载并安装sealos, sealos是个golang的二进制工具，直接下载拷贝到bin目录即可, release页面也可下载$ wget -c https://sealyun.oss-cn-beijing.aliyuncs.com/latest/sealos && \ chmod +x sealos && mv sealos /usr/bin # 下载离线资源包$ wget -c https://seal

监控redis_cluster。

nginx安装#创建目录mkdir -p /data/nginx/{conf,log,conf/conf.d}#创建nginx配置nginx.confcat > /data/nginx/conf/nginx.conf << 'EOF'user nginx;worker_processes auto;error_log /var/log/nginx/error.log notice;pid /var/run/nginx.pid;events

mkdir /data/etcd/{data,bin} -pcurl -L https://github.com/etcd-io/etcd/releases/download/v3.5.0/etcd-v3.5.0-linux-amd64.tar.gz -o /tmp/etcd-v3.5.0-linux-amd64.tar.gztar zxvf /tmp/etcd-v3.5.0-linux-amd64.tar.gz -C /data/etcdcd /data/etcd/etcd-v3.5.0-linu

在docker-compose.yml文件中里添加下面内容： extra_hosts: - "www.163.com:172.17.3.1"

Zookeeper环境准备主机ip消息端口通信端口节点目录192.168.11.21121812888:3888zookeeper192.168.11.21221812888:3888zookeeper192.168.11.21321812888:3888zookeeperActiveMQ环境准备：主机IP集群通信端口消息端口控制台端口节点目录192.168.11.2116262161616，61613，5672，

1、安装zookeeperIP角色备注192.168.11.192server节点1192.168.11.193server节点2192.168.11.194server,client,ui节点32、配置文件说明zoo.confauthProvider.1=org.apache.zookeeper.server.auth.SASLAuthenticationProviderrequireClientAuthScheme=sasljaasLogi

mkdir -p /data/openldap/{data,config,init,ldap-account-manager,self-service-password}cd /data/openldap/vi docker-compose.yamlversion: "3"services: ldap: container_name: "ldap" hostname: ldap2.fly.cn image: "osixia/openldap:latest" re

1、 docker in docker(dind)有时需要在容器内执行 docker 命令，比如：在 jenkins 容器内运行 docker 命令执行构建镜像直接在 docker 容器内嵌套安装 docker 未免太过臃肿更好的办法是：容器内仅部署 docker 命令行工具（作为客户端），实际执行交由宿主机内的 docker-engine（服务器）2、 两种方式在docker容器内启动一个docker daemon，对外提供服务。每个运行中的容器，都是一个进程，这个进程都托管在docker d

version: '3'services: gitlab: container_name: "gitlab" image: 'gitlab/gitlab-ce:13.10.2-ce.0' restart: unless-stopped hostname: 'gitlab.fly.cn' environment: TZ: 'Asia/Shanghai' GITLAB_OMNIBUS_CONFIG: |

生产环境中grafana的数据在做持久化,以grafana-piechart-panel为例如：下载离线插件wget -nv https://grafana.com/api/plugins/grafana-piechart-panel/versions/latest/download -O /tmp/grafana-piechart-panel.zip把插件copy进容器中kubectl cp /tmp/grafana-piechart-panel.zip -n monitor prom

yum install rpcbind nfs-utils -y systemctl enable nfs-server.servicesystemctl enable rpcbind.servicesystemctl start nfs-server.servicesystemctl start rpcbind.servicemkdir /nfsdataecho '/nfsdata 192.168.11.0/24(rw,sync,fsid=0)' >> /etc/export

helm 3kubernetes 1.16+helm repo add prometheus-community https://prometheus-community.github.io/helm-chartshelm search repo prometheushelm pull prometheus-community/kube-prometheus-stackkubectl create ns monitoringhelm install prometheus kube-prometh

#ceph clusterwget https://raw.githubusercontent.com/rook/rook/release-1.6/cluster/examples/kubernetes/ceph/crds.yamlwget https://raw.githubusercontent.com/rook/rook/release-1.6/cluster/examples/kubernetes/ceph/common.yamlwget https://raw.githubuserconte

wget https://download.qemu.org/qemu-5.2.0.tar.xz1、binfmt支持#重置binfmtdocker run --rm --privileged multiarch/qemu-user-static --reset -p yes #设置binfmtdocker run --rm --privileged multiarch/qemu-user-static -p yes 确认是否成功[root@localhost binfmt_misc]#

kind: IngressapiVersion: extensions/v1beta1metadata: name: dashboard namespace: kube-system annotations: kubernetes.io/ingress.class: traefik #ingress.kubernetes.io/ssl-passthrough: "true"spec: rules: - host: dashboard.digitalgd.com http:

ingress --traefikapiVersion: networking.k8s.io/v1beta1kind: Ingressmetadata: name: test-ingress namespace: test labels: app: test-cors-app annotations: kubernetes.io/ingress.class: traefik traefik.ingress.kubernetes.io/custom-respons

编辑configmap (coredns)kubectl edit configmap -n kube-system coredns 增加hosts节点 Corefile: | .:53 { errors health { lameduck 5s } ready kubernetes cluster.local in-addr.arpa ip6.arpa { po

1、docker-entrypoint.sh#!/bin/bashset -eif [[ $redis_url ]]; then sed -i 's#redis://127.0.0.1:6379#redis://'${redis_url}'#' /data/smartgate/etc/settings.jsonfiif [[ $redis_password ]]; then sed -i 's#password=password#password='${redis_

清理docker0、flannel、kube-ipvs0、dummy0网络yum -y install bridge-utilsifconfig docker0 downbrctl delbr docker0ip link del flannel.1ip link del kube-ipvs0ip link del dummy0清理所有docker运行镜像docker ps -a|awk '{print $1}'|grep -v CONTAINER|xargs docker r.