Question

在部署非常简单的Pod时，我看不到任何日志输出：

myconfig.yaml：

apiVersion: v1
kind: Pod
metadata:
  name: counter
spec:
  containers:
  - name: count
    image: busybox
    args: [/bin/sh, -c,
            'i=0; while true; do echo "$i: $(date)"; i=$((i+1)); sleep 1; done']

然后

kubectl apply -f myconfig.yaml

这摘自以下官方教程：https://kubernetes.io/docs/concepts/cluster-administration/logging/#basic-logging-in-kubernetes

吊舱运行正常：

kubectl describe pod counter
Name:         counter
Namespace:    default
Node:         ip-10-0-0-43.ec2.internal/10.0.0.43
Start Time:   Tue, 20 Nov 2018 12:05:07 -0500
Labels:       <none>
Annotations:  kubectl.kubernetes.io/last-applied-configuration={"apiVersion":"v1","kind":"Pod","metadata":{"annotations":{},"name":"counter","namespace":"default"},"spec":{"containers":[{"args":["/bin/sh","-c","i=0...
Status:       Running
IP:           10.0.0.81
Containers:
  count:
    Container ID:  docker://d2dfdb8644b5a6488d9d324c8c8c2d4637a460693012f35a14cfa135ab628303
    Image:         busybox
    Image ID:      docker-pullable://busybox@sha256:2a03a6059f21e150ae84b0973863609494aad70f0a80eaeb64bddd8d92465812
    Port:          <none>
    Host Port:     <none>
    Args:
      /bin/sh
      -c
      i=0; while true; do echo "$i: $(date)"; i=$((i+1)); sleep 1; done
    State:          Running
      Started:      Tue, 20 Nov 2018 12:05:08 -0500
    Ready:          True
    Restart Count:  0
    Environment:    <none>
    Mounts:
      /var/run/secrets/kubernetes.io/serviceaccount from default-token-r6tr6 (ro)
Conditions:
  Type           Status
  Initialized    True 
  Ready          True 
  PodScheduled   True 
Volumes:
  default-token-r6tr6:
    Type:        Secret (a volume populated by a Secret)
    SecretName:  default-token-r6tr6
    Optional:    false
QoS Class:       BestEffort
Node-Selectors:  <none>
Tolerations:     node.kubernetes.io/not-ready:NoExecute for 300s
                 node.kubernetes.io/unreachable:NoExecute for 300s
Events:
  Type    Reason                 Age   From                                Message
  ----    ------                 ----  ----                                -------
  Normal  Scheduled              16m   default-scheduler                   Successfully assigned counter to ip-10-0-0-43.ec2.internal
  Normal  SuccessfulMountVolume  16m   kubelet, ip-10-0-0-43.ec2.internal  MountVolume.SetUp succeeded for volume "default-token-r6tr6"
  Normal  Pulling                16m   kubelet, ip-10-0-0-43.ec2.internal  pulling image "busybox"
  Normal  Pulled                 16m   kubelet, ip-10-0-0-43.ec2.internal  Successfully pulled image "busybox"
  Normal  Created                16m   kubelet, ip-10-0-0-43.ec2.internal  Created container
  Normal  Started                16m   kubelet, ip-10-0-0-43.ec2.internal  Started container

运行时不显示任何内容

kubectl logs counter --follow=true

Answer 1

我唯一想到的可能是导致它不输出日志的原因是，如果您在/etc/docker/docker.json配置文件中为运行Pod的节点配置了default logging driver for Docker：

{
  "log-driver": "anything-but-json-file",
}

从本质上讲，这将使Docker不再为kubectl logs <podid> -c <containerid>之类的东西输出stdout / stderr日志。您可以查看节点（10.0.0.43）中的pod的容器中配置了什么：

$ docker inspect -f '{{.HostConfig.LogConfig.Type}}' <container-id>

Answer 2

我遵循了Seenickode的评论，并使其正常运行。

我发现1.10.11或1.11.5（aws中的当前版本）的新cloudformation模板与我的堆栈比较很有用。

这是我学到的东西：

允许从群集安全组到辅助节点的端口1025-65535。
允许端口443从控制平面到辅助节点的出口。

然后，kubectl日志开始工作。

此处示例Cloudformation模板更新：

  NodeSecurityGroupFromControlPlaneIngress:
    Type: AWS::EC2::SecurityGroupIngress
    DependsOn: NodeSecurityGroup
    Properties:
      Description: Allow worker Kubelets and pods to receive communication from the cluster control plane
      GroupId: !Ref NodeSecurityGroup
      SourceSecurityGroupId: !Ref ControlPlaneSecurityGroup
      IpProtocol: tcp
      FromPort: 1025
      ToPort: 65535

也

  ControlPlaneEgressToNodeSecurityGroupOn443:
    Type: AWS::EC2::SecurityGroupEgress
    DependsOn: NodeSecurityGroup
    Properties:
      Description: Allow the cluster control plane to communicate with pods running extension API servers on port 443
      GroupId:
        Ref: ControlPlaneSecurityGroup
      DestinationSecurityGroupId:
        Ref: NodeSecurityGroup
      IpProtocol: tcp
      FromPort: 443
      ToPort: 443

Answer 3

使用此：

$ kubectl logs -f counter --namespace default

Answer 4

您在注释中提到的错误表明您的kubelet进程未运行或正在重新启动。

ss -tnpl |grep 10250
LISTEN     0      128         :::10250                   :::*                   users:(("kubelet",pid=1102,fd=21))

检查以上命令，看看pid是否在某个时间间隔内连续变化。

此外，检查/ var / log / messages是否存在任何与节点相关的问题。希望这会有所帮助。

Answer 5

我发现了问题。此处的AWS教程docs.aws.amazon.com/eks/latest/userguide/getting-started.html引用了CloudFormation模板，这些模板未能设置所需的安全组，因此无法正确查看日志。我基本上为k8s工作者节点（EC2实例）打开了所有流量和端口，然后一切正常。

在正式的Kubernetes示例之后，``kubectl日志计数器''未显示任何输出

5 个答案: