K8S容器持续Terminating无法正常关闭(sider-car容器异常,微服务容器正常)

问题

K8S上出现大量持续terminating的Pod,无法通过常规命令删除。需要编写脚本批量强制删除持续temminating的Pod:contribution-xxxxxxx。

解决

  1. 获取terminating状态的pod名称的命令:
bash 复制代码
# 获取media命名空间下,名称带contribution,运行状态除了正常Runing的Pod信息,并只打印第一列Pod名称。
kubectl get pod -n media |grep contribution|grep -v Running|awk '{print $1}'
  1. 强制删除pod的命令:
bash 复制代码
# 强制删除media命令空间下,名为 contribution-7d57bdb6d7-krksf的Pod。
kubectl delete pods contribution-7d57bdb6d7-krksf --grace-period=0 --force -n media
  1. 编写脚本对大量terminating的同名容器强制删除:
bash 复制代码
vim 72_ForceDeleteTerminatingPods.sh

脚本具体内容如下:

bash 复制代码
#!/bin/bash
for POD in `kubectl get pod -n media |grep contribution|grep -v Running|awk '{print $1}'`
do
    kubectl delete pods ${POD} --grace-period=0 --force -n media
done
  1. 执行结果
bash 复制代码
[sysma@prod-k8s-0001 ~]$ chmod +x 72_ForceDeleteTerminatingPods.sh 
[sysma@prod-k8s-0001 ~]$ ./72_ForceDeleteTerminatingPods.sh 
Warning: Immediate deletion does not wait for confirmation that the running resource has been terminated. The resource may continue to run on the cluster indefinitely.
pod "contribution-56695dbc4-dq2bc" force deleted
Warning: Immediate deletion does not wait for confirmation that the running resource has been terminated. The resource may continue to run on the cluster indefinitely.
pod "contribution-58c5f57b54-wkkfs" force deleted
Warning: Immediate deletion does not wait for confirmation that the running resource has been terminated. The resource may continue to run on the cluster indefinitely.
pod "contribution-5f646dd579-tjcf2" force deleted
Warning: Immediate deletion does not wait for confirmation that the running resource has been terminated. The resource may continue to run on the cluster indefinitely.
pod "contribution-7499fc9b64-2v8tm" force deleted
Warning: Immediate deletion does not wait for confirmation that the running resource has been terminated. The resource may continue to run on the cluster indefinitely.
pod "contribution-75d5f4c94c-q42hq" force deleted
Warning: Immediate deletion does not wait for confirmation that the running resource has been terminated. The resource may continue to run on the cluster indefinitely.
pod "contribution-76d45d4bcb-5wtsr" force deleted
Warning: Immediate deletion does not wait for confirmation that the running resource has been terminated. The resource may continue to run on the cluster indefinitely.
pod "contribution-784b5569b7-ggdx9" force deleted
Warning: Immediate deletion does not wait for confirmation that the running resource has been terminated. The resource may continue to run on the cluster indefinitely.
pod "contribution-7d57bdb6d7-krksf" force deleted
Warning: Immediate deletion does not wait for confirmation that the running resource has been terminated. The resource may continue to run on the cluster indefinitely.
pod "contribution-7d57bdb6d7-qjn4w" force deleted
Warning: Immediate deletion does not wait for confirmation that the running resource has been terminated. The resource may continue to run on the cluster indefinitely.
pod "contribution-7ff99c7d9d-8p6bf" force deleted
Warning: Immediate deletion does not wait for confirmation that the running resource has been terminated. The resource may continue to run on the cluster indefinitely.
pod "contribution-7ff99c7d9d-vfg72" force deleted
Warning: Immediate deletion does not wait for confirmation that the running resource has been terminated. The resource may continue to run on the cluster indefinitely.
pod "contribution-b69bb846d-trwx2" force deleted

参考截图


结束

相关推荐
lijun_xiao200923 分钟前
DevOps(devops/k8s/docker/Linux)学习笔记-4
docker·kubernetes·devops
再难也得平24 分钟前
微服务拆分之SpringCloud
java·spring cloud·微服务
坚持的小马2 小时前
k8s中执行脚本如果提示etcd找不到,可以做如下的操作
容器·kubernetes·etcd
easy_coder2 小时前
从HDFS NN报错看Flink+K8s+HDFS:基础、架构与问题关联
hdfs·flink·kubernetes
回忆是昨天里的海3 小时前
k8s-部署springboot容器化应用
java·容器·kubernetes
立早正文3 小时前
Docker从零到一部署DNMP+Redis《全程干货》
docker·容器·php
sas-soft3 小时前
短任务占资源、低频 API 耗成本?混合架构方案:微服务扛核心,Serverless 接弹性需求
微服务·架构·serverless
hkNaruto3 小时前
【k8s】Kubernetes 资源限制设置规范手册 MB与MiB的概念混淆问题
云原生·容器·kubernetes
Mr. Cao code5 小时前
实战:Docker构建Haproxy负载均衡镜像
linux·运维·ubuntu·docker·容器·负载均衡
Xander W5 小时前
基于K8s集群的PyTorch DDP 框架分布式训练测试(开发机版)
人工智能·pytorch·分布式·python·深度学习·kubernetes