无法启动Skydns [英] Cannot start skydns

问题描述

我正在尝试在ubutnu上安装kubernetes(裸机安装). 我安装了一个主节点和一个节点，看起来不错. 我安装了ui插件没有问题，但是当我尝试安装dns插件时，pod会自动重启.

I'm trying to install kubernetes on ubutnu (bare-metal installation). I install one master and one node and it looks fine. I installed the ui addon without a problem but when I tried to install the dns addon the pod restart itself continuously.

在kube-apiserver日志中，我得到了:

in kube-apiserver logs I get:

E1218 12:56:15.298118       5 handlers.go:37] Unable to authenticate the request due to an error: crypto/rsa: verification error
I1218 12:56:15.298220       5 handlers.go:131] GET /api/v1/services: (534.467µs) 401 [[kube2sky/v0.20.2 (linux/amd64) kubernetes/unknown] 172.27.35.39:33013]
E1218 12:56:15.298396       5 handlers.go:37] Unable to authenticate the request due to an error: crypto/rsa: verification error
I1218 12:56:15.298469       5 handlers.go:131] GET /api/v1/endpoints: (493.5µs) 401 [[kube2sky/v0.20.2 (linux/amd64) kubernetes/unknown] 172.27.35.39:33014]
I1218 12:56:16.001321       5 handlers.go:131] GET /healthz: (83.326µs) 0 [[Go 1.1 package http] 127.0.0.1:42096]
E1218 12:56:16.303274       5 handlers.go:37] Unable to authenticate the request due to an error: crypto/rsa: verification error
E1218 12:56:16.303274       5 handlers.go:37] Unable to authenticate the request due to an error: crypto/rsa: verification error

在kube2sky容器日志中，出现以下错误:

And in the kube2sky container log I get the following errors:

E1218 12:57:51.713651       1 reflector.go:136] Failed to list   *api.Service: the server has asked for the client to provide credentials (get services)
E1218 12:57:51.713850       1 reflector.go:136] Failed to list *api.Endpoints: the server has asked for the client to provide credentials (get endpoints)

在我看来，我的服务帐户和令牌有问题，但是我查看了kube2sky实例，发现在/tmp/secrets/kubernetes.io/serviceaccount目录中有一个令牌，而ca.crt是群集的正确ca

looks to me like I have a problem with service account and token but I looked in the kube2sky instance and I see that I have a token in /tmp/secrets/kubernetes.io/serviceaccount directory and the ca.crt is the correct ca of the cluster

我正在使用: Ubuntu的04.04 kubernetes 1.1.2 我尝试从kubernetes 1.1.2的插件目录创建dns插件

I am using: ubuntu 14.04 kubernetes 1.1.2 add I try to create the dns addon from the addon directory of kubernetes 1.1.2

推荐答案

更新api服务器的证书和密钥后，我遇到了同样的问题.删除秘密后，我修复了DNS服务:

I had this same problem after updating the certificate and key of the api-server. I fixed the DNS service after removing the secret:

# Search the secret name
kubectl get pod --namespace=kube-system -l k8s-app=kube-dns -o yaml | grep -A1 serviceaccount

# Delete the current secret
kubectl delete secret/<name-of-the-secret> --namespace=kube-system

我还需要重新启动整个POD，不知道是否有更好的方法:

I also need to restart the whole POD, don't know if there is a better approach:

kubectl delete svc/kube-dns rc/kube-dns-v9 --namespace=kube-system
kubectl create -f dns-addon.yaml 

这篇关于无法启动Skydns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！