hknaruto的专栏

但是通过add-node、add-master方法，add-master会覆盖node角色，导致节点只是master，如果只有master，那么这个集群对业务无法使用。按照前面的思路，提前解决遇到的问题，然后基于hosts模板文件，部署3个即是master又是node的节点（不嫌麻烦的化，可以搞6个虚拟机，3个master 3个node 参考第【2】篇即可完成）destroy并重启三个节点，删除cluster目录，重新来一遍。根据千米的经验，kubeasz部署遇到的问题需要的解决方案有。

接上篇为排除机械硬盘速度慢影响，将三台虚拟机移动到nvme硬盘上，重新setup，仍然报错TASK [kube-node : 轮询等待kubelet启动] *********************************************************************************FAILED - RETRYING: [10.47.76.74]: 轮询等待node达到Ready状态 (8 retries left).

docker-24.0.7.tgz 100%[=====================================================================================>] 66.60M 1.54MB/s 用时 41s。正在连接 mirrors.tuna.tsinghua.edu.cn (mirrors.tuna.tsinghua.edu.cn)|101.6.15.130|:443... 已连接。

livenessProbe: initialDelaySeconds: 60 periodSeconds: 5 timeoutSeconds: 3 successThreshold: 1 failureThreshold: 3 exec: command: - /bin/sh - -c ...

过程 登录对应节点查看pod状态kubectl describe pod -n kube-system kube-flannel-ds-arm64-2xmrg其中有Last State: Terminated Reason: OOMKilled Exit Code: 137 Started: Fri, 26 Nov 2021 10:57:14 +0800 Finished: Fri, 26 N..

标记节点不可部署 kubectl cordon node# kubectl get nodesNAME STATUS ROLES AGE VERSION10.0.0.119 Ready,SchedulingDisabled master 18m v1.18.610.0.0.140 Ready,SchedulingDisabled master 18m v1.18.610.0.0.187

标记污点kubectl taint nodes 192.168.110.106 192.168.110.188 192.168.110.194 glusterfs:NoSchedulenode/192.168.110.106 taintednode/192.168.110.188 taintednode/192.168.110.194 tainted取消污点kubectl taint nodes 192.168.110.106 192.168.110.188 192.168.110.

防火墙开启状态，harbor pod出错Events: Type Reason Age From Message ---- ------ ---- ---- ------- Warning Unhealthy 19m (x319 over 7h28m) kubelet, 10.51.72.167

kubectl top nodes# kubectl top nodeNAME CPU(cores) CPU% MEMORY(bytes) MEMORY%kube-compute-node-1 337m 2% 15922Mi 25%kube-compute-node-2 188m 1% 14327Mi 23%kube-compute-no...

安装部署nfs1. nfs服务器创建目录[yeqiang@harbor ~]$ sudo mkdir /home/nfs/jenkins -p2. 设置其他用户可以写入该目录[yeqiang@harbor ~]$ sudo chmod o+w /home/nfs/jenkins/ -R3. 配置/etc/exports/home/nfs/jenkins 10.51.72.167(rw,sync,no_subtree_check,no_root_squash)4. 启动nf

故障现象bash-4.3# curl www.baidu.comcurl: (6) Couldn't resolve host 'www.baidu.com'dns配置bash-4.3# cat /etc/resolv.conf nameserver 10.68.0.2search jenkins.svc.cluster.local. svc.cluster.local. cluster.local. greatwall.com.cn hknaruto.comoptions ndot

Fedora 31 自带了nfs服务[yeqiang@harbor ~]$ rpm -qi nfs-utilsName : nfs-utilsEpoch : 1Version : 2.4.1Release : 1.rc1.fc31Architecture: x86_64Install Date: Thu 24 Oct 2019 07:13:57 AM CSTGroup : UnspecifiedSize : 16902

1. Docker部署[yeqiang@harbor wrk]$ docker run --name=test -d -p 81:80 nginx:alpine3a37a184cb67d5ea9f608ae1b5c196c679b263872a8edc598d80929d05c7f9d0wrk测试[yeqiang@harbor wrk]$ ./wrk -t4 -c160 -d30s http://localhost:81/Running 30s test @ http://localh

namespace=kube-systemingress-nginx-controller错误日志W0805 02:43:15.093543 6 controller.go:1133] Unexpected error validating SSL certificate "harbor/hknaruto.com" for server "core.harbor.hknaruto.com": x509: certificate is valid for *.hknaruto.com,

生成自签名证书参考https://hknaruto.blog.csdn.net/article/details/79556245得到密钥及证书文件：hknaruto.com.key， hknaruto.com.pem创建k8s secret[yeqiang@localhost openssl-CA]$ kubectl create secret tls hknaruto.com --cert=hknaruto.com.pem --key=hknaruto.com.key -n ..