本文记录了在安装k8s v1.13.2过程中出现的各式问题以及解决方案，不定期更新，供日后查看。正常安装步骤见：
Kubernetes实践指南：kubeadm安装集群K8s:v1.13.2 - 简书
1.kubelet启动报错：W0203 MemoryAccounting CPUAccounting not enabled for pid...
[root@k8s-node2 ~]# service kubelet status
Redirecting to /bin/systemctl status kubelet.service
kubelet.service - kubelet: The Kubernetes Node Agent
Loaded: loaded (/etc/systemd/system/kubelet.service; enabled; vendor preset: disabled)
Drop-In: /etc/systemd/system/kubelet.service.d
└─10-kubeadm.conf
Active: active (running) since 日 2019-02-03 11:35:52 CST; 1h 49min ago
Docs: https://kubernetes.io/docs/
Main PID: 9766 (kubelet)
CGroup: /system.slice/kubelet.service
└─9766 /usr/bin/kubelet --bootstrap-kubeconfig=/etc/kubernetes/bootstrap-kubelet.conf --kubeconfig=/etc/kubernetes/kubelet.conf --config=/var/lib/kubelet/con...
2月 03 13:10:53 k8s-node2 kubelet[9766]: W0203 13:10:53.182621 9766 container_manager_linux.go:804] CPUAccounting not enabled for pid: 9766
2月 03 13:10:53 k8s-node2 kubelet[9766]: W0203 13:10:53.182630 9766 container_manager_linux.go:807] MemoryAccounting not enabled for pid: 9766
2月 03 13:15:53 k8s-node2 kubelet[9766]: W0203 13:15:53.183017 9766 container_manager_linux.go:804] CPUAccounting not enabled for pid: 9085
2月 03 13:15:53 k8s-node2 kubelet[9766]: W0203 13:15:53.183056 9766 container_manager_linux.go:807] MemoryAccounting not enabled for pid: 9085
2月 03 13:15:53 k8s-node2 kubelet[9766]: W0203 13:15:53.183156 9766 container_manager_linux.go:804] CPUAccounting not enabled for pid: 9766
2月 03 13:15:53 k8s-node2 kubelet[9766]: W0203 13:15:53.183161 9766 container_manager_linux.go:807] MemoryAccounting not enabled for pid: 9766
2月 03 13:20:53 k8s-node2 kubelet[9766]: W0203 13:20:53.184116 9766 container_manager_linux.go:804] CPUAccounting not enabled for pid: 9085
2月 03 13:20:53 k8s-node2 kubelet[9766]: W0203 13:20:53.184155 9766 container_manager_linux.go:807] MemoryAccounting not enabled for pid: 9085
2月 03 13:20:53 k8s-node2 kubelet[9766]: W0203 13:20:53.184237 9766 container_manager_linux.go:804] CPUAccounting not enabled for pid: 9766
2月 03 13:20:53 k8s-node2 kubelet[9766]: W0203 13:20:53.184243 9766 container_manager_linux.go:807] MemoryAccounting not enabled for pid: 9766
首先查看内存的使用情况：# free -h 发现并没有存在内存不够的情况。解决办法：增加一个配置文件，明确启用DefaultCPUAccounting和DefaultMemoryAccounting：
# mkdir -p /etc/systemd/system.conf.d
# cat <<EOF >/etc/systemd/system.conf.d/kubernetes-accounting.conf
[Manager]
DefaultCPUAccounting=yes
DefaultMemoryAccounting=yes
EOF
# systemctl daemon-reload && systemctl restart kubelet
2.Kubernetes Node节点执行 kubectl get all 错误：The connection to the server localhost:8080 was refused.
[root@k8s-node2 ~]# kubectl get all
The connection to the server localhost:8080 was refused - did you specify the right host or port?
使用 netstat -ntlp 命令检查是否监听了localhost:8080端口，发现并没有。而在Master节点上使用kubectl命令虽然不会报错，但其8080端口仍然未被监听。
事实上，kubectl命令是通过kube-apiserver接口进行集群管理。该命令可以在Master节点上运行是因为kube-apiserver处于工作状态：
[root@k8s-master ~]# docker ps | grep apiserver
269a09fc31ce 177db4b8e93a "kube-apiserver --..." 20 hours ago Up 20 hours k8s_kube-apiserver_kube-apiserver-k8s-master_kube-system_e65c58fe4249c7d1554ca017bda21943_0
dcf07ff997a1 k8s.gcr.io/pause:3.1 "/pause" 20 hours ago Up 20 hours k8s_POD_kube-apiserver-k8s-master_kube-system_e65c58fe4249c7d1554ca017bda21943_0
而同时，在Node节点上只有kube-proxy和kubelet处于工作状态：
[root@k8s-node1 ~]# docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
fa14d993436a 142953928206 "/install-cni.sh" 20 hours ago Up 20 hours k8s_install-cni_calico-node-clc9p_kube-system_ac5f61a7-26d2-11e9-9274-000c29d747fb_0
4e77ea62ac14 01cfa56edcfc "/usr/local/bin/ku..." 20 hours ago Up 20 hours k8s_kube-proxy_kube-proxy-nzfvg_kube-system_ac5f6294-26d2-11e9-9274-000c29d747fb_0
2bb208e1573d e537e5882f91 "start_runit" 20 hours ago Up 20 hours k8s_calico-node_calico-node-clc9p_kube-system_ac5f61a7-26d2-11e9-9274-000c29d747fb_0
8490970048da k8s.gcr.io/pause:3.1 "/pause" 20 hours ago Up 20 hours k8s_POD_calico-node-clc9p_kube-system_ac5f61a7-26d2-11e9-9274-000c29d747fb_0
f8eb0bb6693b k8s.gcr.io/pause:3.1 "/pause" 20 hours ago Up 20 hours k8s_POD_kube-proxy-nzfvg_kube-system_ac5f6294-26d2-11e9-9274-000c29d747fb_0
因此，kubectl命令其实不是为Node节点的主机准备的，而是应该运行在一个Client主机上：如K8s-Master节点的非root用户。当我们kubeadm init success后，系统会提示我们将admin.conf文件保存到Client主机上：
Your Kubernetes master has initialized successfully!
To start using your cluster, you need to run the following as a regular user:
mkdir -p $HOME/.kube
sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
sudo chown $(id -u):$(id -g) $HOME/.kube/config
You should now deploy a pod network to the cluster.
Run "kubectl apply -f [podnetwork].yaml" with one of the options listed at:
https://kubernetes.io/docs/concepts/cluster-administration/addons/
You can now join any number of machines by running the following on each node
as root:
kubeadm join 192.168.1.120:6443 --token oe50fb.0pt36rwvz2utey4d --discovery-token-ca-cert-hash sha256:60bd336002b8f5d269996f1daf324c0a71814d6a25d82ab7b1d17ddeddd68860
我们查看 /etc/kubernetes/admin.conf文件
apiVersion: v1
clusters:
- cluster:
certificate-authority-data: LS0tLS1CRUdJTiBDRVJUSUZJQ0FURS0tLS0tCk1JSUN5RENDQWJDZ0F3SUJBZ0lCQURBTkJna3Foa2lHOXcwQkFRc0ZBREFWTVJNd0VRWURWUVFERXdwcmRXSmwKY201bGRHVnpNQjR
YRFRFNU1ESXdNakE1TlRZMU5sb1hEVEk1TURFek1EQTVOVFkxTmxvd0ZURVRNQkVHQTFVRQpBeE1LYTNWaVpYSnVaWFJsY3pDQ0FTSXdEUVlKS29aSWh2Y05BUUVCQlFBRGdnRVBBRENDQVFvQ2dnRUJBTHFwCnRScTl4Smk5cz
NTdUVsVXljNmMwcGhhWSs4OHlQcUpYQnBsZk1YOFpJcmJVWDdHTFB5ZDVzZlBrS0lrblJ6dUgKeTZxb091NUVVbWtYZ1dldlNzK1JITGdYbHNuUFBhSHhCK0o5Y1pxNjg5cnQrd3huMDl6OVpNT0ROc0ZMTHRVMgoxUEFoY3lRZ
TNOZVBPSUdseHQvckZRRlBUV05KQTErbmJCSk9sZEhlVUhmWjNaaVcwbVFHM0IrWk1SUUpWdkM0CmIrdHRVaUpaK3FQL09SaUZKR3VUYmJzS2tsUlNIaG9xMnVtSExxYmhLTVJNQXRRbTIxZWMzaXVxVVp4QWl4MlcKdnR1Uzgr
ZUV0U3lIQW8xTm00bzd2dFh3eGVrTkYzT2lVOUZ5T1VvS3NxdVRKenVhdk9UdVJoYjd1REpQZERoaApFRzZzMlZvUjZyRDB2UjFmZUZVQ0F3RUFBYU1qTUNFd0RnWURWUjBQQVFIL0JBUURBZ0trTUE4R0ExVWRFd0VCCi93UUZ
NQU1CQWY4d0RRWUpLb1pJaHZjTkFRRUxCUUFEZ2dFQkFBSytla09IT1dQTGhsVzJva2g0bTlRNTRJY3oKOEJPU1VEYnJsSk9iSXFUaWNvWktsOGNNMjM3OTlDcXUrVDh2WHA3YXRQc0xtd2xRK2VVK2lUVUNZVGk3d013Lwo1M1
lxWjNCSHVQS2F0RDNoVGpFRlVIbzFZVHMyYmZqVHZ5Z2hLbGhDVnBGL1k4NmFHOVFUVUxmc0g5VXpwbWtjCk5DZzU3T0tUWjFNc3FQUmIrM1hRSEFCWHVaR1RNVG4zaGVZR2dnYklVaC9vdTJyM2RhdFY0ZWdTaDhveFBJcmoKa
FdhU0JOcmVaaE45a1VsVmNoT3RsZ2lvcDJzR1A0V2RLQisxc2kxU2x2YUI5aGR6VklpTHFGWnlhY3I5ZUlvaAp1ckVib2lZYXovU2hGeSs1UCs1SWViZ0h5QWtuWm5EbXFKT3ZXbjducUNhc3RmYi81bERHYVZCcmxtZz0KLS0t
LS1FTkQgQ0VSVElGSUNBVEUtLS0tLQo=
server: https://192.168.1.120:6443
name: kubernetes
即可发现，当Client使用该config文件启动kubelet后，他将访问Master节点的6443端口获得数据（Master 6443端口是处于LISTEN状态的），而非localhost:8080端口（因为Node节点无法找到该config文件）。我们也可以把Client客户端放在其他主机中，甚至Node节点。只要将该config文件按照系统提示方式添加到Client客户端中即可。我们使用scp命令将文件发送至目标主机：
# scp -r .kube/ 192.168.1.110:/root //在此我直接将/root/.kube文件夹发至目标主机
即可实现使用kubectl访问Master节点。
[root@localhost .kube]# kubectl get no
NAME STATUS ROLES AGE VERSION
k8s-master Ready master 18h v1.13.2
k8s-node1 Ready <none> 18h v1.13.2
k8s-node2 Ready <none> 18h v1.13.2
也就是说，我们正常向Master注册pod的过程也是在Client客户端完成的，而非在Node节点或Master节点完成。
3.kubelet启动报错：E0208 node "k8s-master" not found
2月 08 15:55:36 k8s-master kubelet[6164]: E0208 15:55:36.068126 6164 kubelet.go:2266] node "k8s-master" not found
2月 08 15:55:36 k8s-master kubelet[6164]: E0208 15:55:36.169675 6164 kubelet.go:2266] node "k8s-master" not found
2月 08 15:55:36 k8s-master kubelet[6164]: E0208 15:55:36.238707 6164 kubelet_node_status.go:94] Unable to register node "k8s-master" with API server: Post https://192.168.1.120:6443/api/v1/nodes: dial tcp 192.168.1.120:6443: connect: connection refused
kubeadm在Master节点也安装了kubelet，默认情况下不参与负载。这个错误比较明显，即节点（kubelet）无法连接至Master（kube-apiserver)，不是因为Master节点6443端口关闭，而是后来我修改过Master节点的IP地址，导致旧的IP地址无法正确匹配。解决方法其一可以通过kubeadm reset 重新安装。这里我们尝试修改已安装好的kubernetes参数来使其正确运行。
[root@k8s-master ~]# cd /etc/kubernetes && ls
总用量 36
-rw------- 1 root root 5455 2月 8 16:05 admin.conf
-rw------- 1 root root 5487 2月 8 16:05 controller-manager.conf
-rw------- 1 root root 5483 2月 8 16:06 kubelet.conf
drwxr-xr-x. 2 root root 113 2月 8 16:08 manifests
drwxr-xr-x. 3 root root 4096 2月 2 17:56 pki
-rw------- 1 root root 5435 2月 8 16:08 scheduler.conf
将conf文件中旧的IP地址（192.168.1.120）修改为（192.168.111.120）并保存，重新加载kubelet服务（文件夹内也有conf文件）
[root@k8s-master ~]# systemctl daemon-reload
[root@k8s-master ~]# systemctl restart kubelet && journactl -xefu kubelet
2月 08 16:47:04 k8s-master kubelet[19409]: E0208 16:47:04.188505 19409 kubelet.go:2266] node "k8s-master" not found
2月 08 16:47:04 k8s-master kubelet[19409]: E0208 16:47:04.290432 19409 kubelet.go:2266] node "k8s-master" not found
2月 08 16:47:04 k8s-master kubelet[19409]: E0208 16:47:04.326230 19409 reflector.go:134] k8s.io/kubernetes/pkg/kubelet/kubelet.go:444: Failed to list *v1.Service: Get https://192.168.111.120:6443/api/v1/services?limit=500&resourceVersion=0: x509: certificate is valid for 10.96.0.1, 192.168.1.120, not 192.168.111.120
2月 08 16:47:04 k8s-master kubelet[19409]: E0208 16:47:04.356546 19409 reflector.go:134] k8s.io/kubernetes/pkg/kubelet/config/apiserver.go:47: Failed to list *v1.Pod: Get https://192.168.111.120:6443/api/v1/pods?fieldSelector=spec.nodeName%3Dk8s-master&limit=500&resourceVersion=0: x509: certificate is valid for 10.96.0.1, 192.168.1.120, not 192.168.111.120
2月 08 16:47:04 k8s-master kubelet[19409]: E0208 16:47:04.362324 19409 reflector.go:134] k8s.io/kubernetes/pkg/kubelet/kubelet.go:453: Failed to list *v1.Node: Get https://192.168.111.120:6443/api/v1/nodes?fieldSelector=metadata.name%3Dk8s-master&limit=500&resourceVersion=0: x509: certificate is valid for 10.96.0.1, 192.168.1.120, not 192.168.111.120
从错误日志判断来看，是Master上的kubelet在与同一节点上的kube-apiserver通信过程中，发现这个apiserver返回的tls证书是属于192.168.1.120的，而非192.168.111.120的apiserver，于是报了错。为了要解决这个问题，我们需要为新的IP地址生成自己的数字证书。贴上参考链接：
生成apiserver数字证书（3.2）
stackoverflow invalid-x509
（好了，这个问题得到此为止了，因为我要自己生成太多的数字证书。最后我在各个节点上重新安装了一遍。）
4.不关闭swap进行安装k8s的解决方案（以kubeadm为例）
1.kubelet启动时加入参数--fail-swap-on=false并重启(KUBELET_EXTRA_ARGS：/etc/sysconfig/kubelet）
2.kubeadm init --ignore-preflight-errors=Swap（在kubeadm join时也需要手动加入--ignore...）。
5.unknown container "/system.slice/kubelet.service"：
kubelet运行时报错：
Failed to get system container stats for "/system.slice/kubelet.service": failed to get cgroup stats for "/system.slice/kubelet.service": failed to get container info for "/system.slice/kubelet.service": unknown container "/system.slice/kubelet.service"：
启动时添加参数：--runtime-cgroups=/systemd/system.slice
--kubelet-cgroups=/systemd/system.slice
