Kubernetes常见错误总结

1、屏幕持续打印Pod日志报error: unexpected EOF错误

Kubernetes: requesting flag for "kubectl logs" to avoid 5-minute timeout if no stdout/stderr 

When running kubectl logs --follow on a pod, after 5 minutes of no stdout/stderr, we received:

$ kubectl --kubeconfig=config --namespace=foo logs --follow foo-oneoff-w8npn --container bar

######################################
#                                    #
#  /system_tests/test_derp.py (1/4)  #
#                                    #
######################################

RESULTS: [/system_tests/test_derp/TestDerp] Ran 4 tests in 130.044 s, ALL TESTS PASSED


######################################
#                                    #
#  /system_tests/test_fuzz.py (2/4)  #
#                                    #
######################################

error: unexpected EOF

 解决方案:

For posterity, the problem here is that we're using haproxy to serve a VIP to balance HA Kubernetes masters.

Specifically we're using HA-Proxy v1.4.21, and we have this in our haproxy cfg:

defaults
  timeout client  500000
  timeout server  500000

2、K8s集群初始化成功后,kubectl get nodes 查看节点信息时报错

报错信息:The connection to the server localhost:8080 was refused - did you specify the right host or port?

解决方法:

mkdir -p $HOME/.kube
sudo cp -i /etc/kubernetes/admin.conf $HOME/.kube/config
sudo chown $(id -u):$(id -g) $HOME/.kube/config

3、master节点的calico组件是0/1 Running状态,其他worker节点的calico组件是1/1 Running状态,describe pod发现是

Readiness probe failed: calico/node is not ready: BIRD is not ready: BGP not established with 10.244.0.1,10.244.2.12020-04-13 06:29:59.582 [INFO][682] health.go 156: Number of node(s) with BGP peering established = 0

解决办法:
修改的calico.yaml,新增两行:

            - name: IP_AUTODETECTION_METHOD
              value: "interface=eth0"

value指向从ip a看到的实际网卡名。结果如下:

            # Cluster type to identify the deployment type
            - name: CLUSTER_TYPE
              value: "k8s,bgp"
            - name: IP_AUTODETECTION_METHOD
              value: "interface=eth0"
            # Auto-detect the BGP IP address.
            - name: IP
              value: "autodetect"
            # Enable IPIP
            - name: CALICO_IPV4POOL_IPIP
              value: "Always"

等一会就正常了。

参考:https://blog.csdn.net/majixiang1996/article/details/105438506/

4、使用kubeadm安装kubernetes集群etcd一直卡在starting状态,查看etcd日志报2020-12-08 17:11:10.741954 I | embed: rejected connection from "192.168.100.179:47288" (error "tls: failed to verify client‘s certificate: x509: certificate has expired or is not yet valid", ServerName "") 错误

错误分析:原因是生成证书的机器时间快于etcd服务器时间,导致服务器验证时,证书超出了时间使用范围。

解决办法:

  1)服务器和生成证书机器进行时间同步更新(高可用k8s集群需要master节点之间时间同步)。

  2)或者直接调整生成证书的机器时间,使其慢于其他节点服务器的时间。

不推荐第二种方式。

参考:http://www.bubuko.com/infodetail-3670080.html

5、如果使用nfs作为存储方案所有节点都需要保证安装好nfs-client

检查服务器是否已安装好nfs-utils、rpcbind、libtirpc包。

参考:http://www.bubuko.com/infodetail-3546759.html

6、报错 cannot allocate memory 或者 no space left on device ,修复K8S内存泄露问题

链接:https://www.cnblogs.com/zhangmingcheng/p/14309962.html

7、执行kubectl命令时报错 error: You must be logged in to the server (Unauthorized)

链接:https://www.cnblogs.com/zhangmingcheng/p/14317551.html

8、kubectl delete pod kube-scheduler-xxxx  -n=kube-system 后,这个pod没有重启直接删掉了,重启kubelet服务使调度器pod重启

9、关于在k8s-v1.20以上版本使用nfs作为storageclass出现selfLink was empty, can‘t make reference错误

问题原因:kubernetes 1.20版本 禁用了 selfLink

解决方法(每个master节点都要执行):

vim /etc/kubernetes/manifests/kube-apiserver.yaml

新增配置: - --feature-gates=RemoveSelfLink=false

由于kube-apiserver是静态Pod,修改完静态Pod声明文件(kube-apiserver.yaml)后,无需apply最新的声明文件(kube-apiserver.yaml),即可自动修改静态Pod配置。静态Pod详细内容见:Kubernetes静态Pod 

10、k8s 集群 metrics-server 组件报 authentication.go:65] Unable to authenticate the request due to an error: x509: certificate has expired or is not yet valid 错误

原因 :k8s集群证书更新后没有自动重启对应的组件

解决方法:

执行如下命令后解决

docker ps |grep -E 'k8s_kube-apiserver|k8s_kube-controller-manager|k8s_kube-scheduler|k8s_etcd_etcd' | awk -F ' ' '{print $1}' |xargs docker restart

链接:https://www.cnblogs.com/ligang0357/p/15561680.html

11、kubelet 报如下错误:failed to run Kubelet: running with swap on is not supported, please disable swap! or set --fail-swap-on flag to false. /proc/swaps contained: [Filename...

原因: 当前机器没有关闭swap,关闭当前机器swap即可。

链接:https://www.cnblogs.com/zhangmingcheng/p/12671749.html

12、使用systemctl cat kubelet命令可以看到通过systemd配置的与kubelet有关的所有配置文件的路径与内容

[root@zmc-manage-uat-106 ~]# systemctl cat kubelet
# /etc/systemd/system/kubelet.service
[Unit]
Description=kubelet: The Kubernetes Node Agent
Documentation=http://kubernetes.io/docs/

[Service]
ExecStart=/usr/local/bin/kubelet
Restart=always
StartLimitInterval=0
RestartSec=10

[Install]
WantedBy=multi-user.target

# /etc/systemd/system/kubelet.service.d/10-kubeadm.conf
# Note: This dropin only works with kubeadm and kubelet v1.11+
[Service]
Environment="KUBELET_KUBECONFIG_ARGS=--bootstrap-kubeconfig=/etc/kubernetes/bootstrap-kubelet.conf --kubeconfig=/etc/kubernetes/kubelet.conf"
Environment="KUBELET_CONFIG_ARGS=--config=/var/lib/kubelet/config.yaml"
# This is a file that "kubeadm init" and "kubeadm join" generate at runtime, populating the KUBELET_KUBEADM_ARGS variable dynamically
EnvironmentFile=-/var/lib/kubelet/kubeadm-flags.env
# This is a file that the user can use for overrides of the kubelet args as a last resort. Preferably, the user should use
# the .NodeRegistration.KubeletExtraArgs object in the configuration files instead. KUBELET_EXTRA_ARGS should be sourced from this file.
EnvironmentFile=-/etc/default/kubelet
Environment="KUBELET_EXTRA_ARGS=--node-ip=10.20.31.106 --hostname-override=zmc-manage-uat-106 "
ExecStart=
ExecStart=/usr/local/bin/kubelet $KUBELET_KUBECONFIG_ARGS $KUBELET_CONFIG_ARGS $KUBELET_KUBEADM_ARGS $KUBELET_EXTRA_ARGS
[root@zmc-manage-uat-106 ~]# 

参考:https://www.cnblogs.com/wangqingyong/p/14542937.html

13、kubectl get cs显示scheduler Unhealthy、controller-manager Unhealthy错误

[root@node1 ~]# kubectl get componentstatuses 
Warning: v1 ComponentStatus is deprecated in v1.19+
NAME                 STATUS      MESSAGE                                                                                       ERROR
scheduler            Unhealthy   Get "http://127.0.0.1:10251/healthz": dial tcp 127.0.0.1:10251: connect: connection refused   
controller-manager   Unhealthy   Get "http://127.0.0.1:10252/healthz": dial tcp 127.0.0.1:10252: connect: connection refused   
etcd-0               Healthy     {"health":"true"}       

错误分析:

1)首先确认没有启动10251、10252端口

 netstat -ntlup

2)确认schedule和controller-manager组件配置是否禁用了非安全端口 

解决方法:

分别修改/etc/kubernetes/manifests/kube-scheduler.yaml和/etc/kubernetes/manifests/kube-controller-manager.yaml配置文件,将容器启动参数中 --port=0注释掉,该--port=0表示禁止使用非安全的http接口。下面以修改调度器配置文件为例,控制器配置文件修改和调度器一致。

containers:
  - command:
    - kube-scheduler
    - --authentication-kubeconfig=/etc/kubernetes/scheduler.conf
    - --authorization-kubeconfig=/etc/kubernetes/scheduler.conf
    - --bind-address=0.0.0.0
    - --feature-gates=ExpandCSIVolumes=true,RotateKubeletServerCertificate=true
    - --kubeconfig=/etc/kubernetes/scheduler.conf
    - --leader-elect=true
  #  - --port=0

重启kubelet服务

systemctl restart kubelet

问题解决:

[root@node1 ~]# kubectl get cs
NAME                 STATUS    MESSAGE             ERROR
controller-manager   Healthy   ok                  
scheduler            Healthy   ok                  
etcd-0               Healthy   {"health":"true"} 

再次通过netstat -ntlup命令验证10251、10252端口是否监听:

.......        
tcp6       0      0 :::10250                :::*                    LISTEN      20137/kubelet       
tcp6       0      0 :::10251                :::*                    LISTEN      13290/kube-schedule 
......

14、configmap中json或者yaml文件内容格式问题

1)问题现象说明:

比如我有一个yaml格式的配置文件config.yaml,内容如下:

service:
  port: 8001
  # release
  mode: release
  url: kube-prometheus.monitoring
  port: 9090
common:
  file:
    upload: /opt/data/common/file/

把该配置文件使用configmap方式创建:

kubectl create configmap myconfig --from-file=config.yaml

我们希望在命令行终端查看到的内容如下:

[root@node1 ~]# kubectl get configmaps myconfig -o yaml
apiVersion: v1
data:
  config.yaml: |
    service:
      port: 8001
      # release
      mode: release
      url: kube-prometheus.monitoring
      port: 9090
    common:
      file:
        upload: /opt/data/common/file/
kind: ConfigMap
metadata:
  creationTimestamp: 2019-04-24T05:59:52Z
  name: myconfig
  namespace: default
  resourceVersion: "1774515"
  selfLink: /api/v1/namespaces/default/configmaps/myconfig
  uid: 2d066fef-6656-11e9-96e0-005056bf291a
[root@node1 ~]# 

但是有时候我们看到如下格式错乱的configmap内容:

[root@intellif-0 ~]# kubectl get configmaps myconfig -o yaml
apiVersion: v1
data:
  config.yaml: "service:\n  port: 8001\n  # release\n  mode: release\n  url: kube-prometheus.monitoring\n
    \ port: 9090\ncommon:\n  file:\n    upload: /opt/data/common/file/  \n"
kind: ConfigMap
metadata:
  creationTimestamp: 2019-04-24T06:01:42Z
  name: myconfig
  namespace: default
  resourceVersion: "1774818"
  selfLink: /api/v1/namespaces/default/configmaps/myconfig
  uid: 6eac963e-6656-11e9-96e0-005056bf291a
[root@intellif-0 ~]# 

2)分析与解决:

问题原因: 文件中某一行结尾有空格,示例中是最后一行(使用vim工具:set invlist
正常的yaml文件

# vim config.yaml
service:$
  port: 8001$
  # release$
  mode: release$
  url: kube-prometheus.monitoring$
  port: 9090$
common:$
  file:$
    upload: /opt/data/common/file/$                                                                                                                                                                                     
~                                                                                                                                                                                                                                                                                                                                                                                 
:set invlist 

异常的yaml文件

# vim config.yaml
service:$
  port: 8001$
  # release$
  mode: release$
  url: kube-prometheus.monitoring$
  port: 9090$
common:$
  file:$
    upload: /opt/data/common/file/  $                                                                                                                                                                                     
~                                                                                                                                                                                                                                                                                                                                                                                
:set invlist 

参考:https://blog.csdn.net/zhangxiangui40542/article/details/89491740

15、kubelet日志报不能检索密钥,导致拉取镜像不成功问题

kubelet日志:

kubelet: W1104 18:34:07.882008     843 kubelet_pods.go:863] Unable to retrieve pull secret pro/gb28181server for pro/xxxserver-v1-76cc9b558d-wmbgf due to secret "gb28181server" not found.  The image pull may not succeed.

问题分析:

检查Pod的Yaml文件,发现配置文件中配置了gb28181server这个镜像拉取密钥,但实际当前Pod所在namespace下面是没有这个密钥的。

apiVersion: v1
kind: Pod
metadata:
 ......
spec:
  .......
  enableServiceLinks: true
  imagePullSecrets:
  - name: gb28181server
  nodeName: 129
  ........

解决方案:

1)如果镜像所在harbor项目是公开项目的话,那么kubelet是有权限直接拉取镜像的,这种情况下可以删掉Pod Yaml文件中的imagePullSecrets配置项;

2)如果镜像所在harbor项目是私有项目的话,那么必须创建认证harbor私有仓库的gb28181server密钥。

参考:kubernetes的imagePullSecrets如何生成及使用 - 走看看

16、Centos7服务器Kubernetes新加入节点访问外部域名不通问题

背景:某项目Kubernetes集群新纳入了12台worker节点,新纳入节点和其他集群节点(master节点和worker节点)网络都是连通的,但是项目实施人员却反馈调度到这12台节点上容器组内部是连不通外部域名的,直接在这12台节点上访问外部域名正常。

问题分析:节点上访问外部域名正常,说明节点dns server配置是没问题的;而容器内部访问不通很可能是以下问题导致:

1)集群dns组件(coredns)异常,但是老节点上的容器组内部是能正常解析外部域名的,所以排除此原因;

2)新加入节点存在未开的端口策略,经检查防火墙策略单及相应测试,也都没问题;

3)容器组配置了错误的 dns server,经测试新节点不仅外部域名不通,k8s内部服务也访问不通,所以基本锁定容器组节点配置了错误的dns server,经检查确实如此。

检查Kubernetes集群DNS服务地址:        

[root@master1 ~]# kubectl get svc -n=kube-system |grep dns
coredns                       ClusterIP   10.234.0.3       <none>        53/UDP,53/TCP,9153/TCP         115d

检查新加入节点容器组DNS配置:

vim /var/lib/kubelet/config.yaml
xxxxx
clusterDNS:
- xxx.xxx.xxx.xxx
clusterDomain: cluster.local
xxxxx

经排查两者dns server不一致。

解决方案:修改新加入节点容器组DNS配置。

vim /var/lib/kubelet/config.yaml
xxxxx
clusterDNS:
- 10.234.0.3
clusterDomain: cluster.local
xxxxx

重启新加节点kubelet服务和容器运行时服务,问题得以解决。

 

posted @ 2020-09-30 11:09  人艰不拆_zmc  阅读(1928)  评论(0编辑  收藏  举报