Kubernetes Pod 在“Pod 沙箱已更改，它将被杀死并重新创建"时失败 [英] Kubernetes pods failing on "Pod sandbox changed, it will be killed and re-created"

问题描述

在 Google Container Engine 集群 (GKE) 上，我有时会看到一个 pod(或更多)未启动并查看其事件，我可以看到以下内容

On a Google Container Engine cluster (GKE), I see sometimes a pod (or more) not starting and looking in its events, I can see the following

Pod 沙箱改变了，它会被杀死并重新创建.

Pod sandbox changed, it will be killed and re-created.

如果我等待 - 它只会不断重试.
如果我删除 pod，并允许部署的副本集重新创建它们，它将正常启动.

If I wait - it just keeps re-trying.
If I delete the pod, and allow them to be recreated by the Deployment's Replica Set, it will start properly.

行为不一致.

Kubernetes 版本 1.7.6 和 1.7.8

Kubernetes versions 1.7.6 and 1.7.8

有什么想法吗?

推荐答案

我可以看到在 Google Cloud 中发布的以下消息状态仪表板:

我们正在调查一个影响 Google Container Engine (GKE) 集群的问题，该问题在 docker 崩溃或在节点上重新启动后，无法安排 Pod.

"We are investigating an issue affecting Google Container Engine (GKE) clusters where after docker crashes or is restarted on a node, pods are unable to be scheduled.

据信该问题会影响所有运行 Kubernetes v1.6.11、v1.7.8 和 v1.8.1 的 GKE 集群.

The issue is believed to be affecting all GKE clusters running Kubernetes v1.6.11, v1.7.8 and v1.8.1.

我们的工程团队建议:如果节点的版本为 v1.6.11，请将您的节点降级到 v1.6.10.如果节点版本为 v1.7.8，请将您的节点降级到 v1.7.6.如果节点在 v1.8.1，请将您的节点降级到 v1.7.6.

Our Engineering Team suggests: If nodes are on release v1.6.11, please downgrade your nodes to v1.6.10. If nodes are on release v1.7.8, please downgrade your nodes to v1.7.6. If nodes are on v1.8.1, please downgrade your nodes to v1.7.6.

工程团队还在本文档.这些变通办法适用于无法降级其节点的客户."

Alternative workarounds are also provided by the Engineering team in this doc . These workarounds are applicable to the customers that are unable to downgrade their nodes."

这篇关于Kubernetes Pod 在“Pod 沙箱已更改，它将被杀死并重新创建"时失败的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！