Kubernetes- 容忍度：node.kubernetes.io/unreachable:NoExecute for 300s

Question

我是 Kubernetes 的新手，并且遇到了一些错误。我想在我的本地系统（mac）上创建 Kubernetes 集群。

我的部署.yaml --

apiVersion: apps/v1
kind: Deployment
metadata:
  name: sv-premier
spec:
  selector:
    matchLabels:
      app: sv-premier
  template:
    metadata:
      labels:
        app: sv-premier
    spec:
      volumes:
      - name: google-cloud-key
        secret:
          secretName: gcp-key
      containers:
      - name: sv-premier
        image: gcr.io/proto/premiercore1:latest
        imagePullPolicy: Always
        command: ["/bin/sh"]
        args: ["-c", "while true; do echo Done Deploying sv-premier; sleep 3600;done"]
        volumeMounts:
        - name: google-cloud-key
          mountPath: /var/secrets/google
        env:
        - name: GOOGLE_APPLICATION_CREDENTIALS
          value: /var/secrets/google/key.json
        ports:
        - containerPort: 8080
      imagePullSecrets:
      - name: imagepullsecretkey

我创建了部署 - kubectl apply -f deployment.yaml

kubectl 获取豆荚

NAME                          READY   STATUS    RESTARTS   AGE
sv-premier-5cc8f599f6-9lrtq   1/1     Running   0          11s

kubectl 描述 pod sv-premier-5cc8f599f6-9lrtq

Name:           sv-premier-5cc8f599f6-9lrtq
Namespace:      default
Priority:       0
Node:           docker-desktop/192.168.65.3
Start Time:     Tue, 11 Feb 2020 19:04:21 +0530
Labels:         app=sv-premier
                pod-template-hash=5cc8f599f6
Annotations:    <none>
Status:         Running
IP:             10.1.0.54
IPs:            <none>
Controlled By:  ReplicaSet/sv-premier-5cc8f599f6
Containers:
  sv-premier:
    Container ID:  docker://b8993b4fc43197947649c7409b37e6d381a8d4cbbe56e550bca83931747ddd3e
    Image:         gcr.io/proto/premiercore1:latest
    Image ID:      docker-pullable://gcr.io/proto/premiercore1@sha256:664778c72c3f79147c4c5b73914292a124009591f479a5e3acf42c444eb62860
    Port:          4343/TCP
    Host Port:     0/TCP
    Command:
      /bin/sh
    Args:
      -c
      while true; do echo Done Deploying sv-premier; sleep 3600;done
    State:          Running
      Started:      Tue, 11 Feb 2020 19:04:24 +0530
    Ready:          True
    Restart Count:  0
    Environment:
      GOOGLE_APPLICATION_CREDENTIALS:  /var/secrets/google/key.json
    Mounts:
      /var/run/secrets/kubernetes.io/serviceaccount from default-token-s4jgd (ro)
      /var/secrets/google from google-cloud-key (rw)
Conditions:
  Type              Status
  Initialized       True 
  Ready             True 
  ContainersReady   True 
  PodScheduled      True 
Volumes:
  google-cloud-key:
    Type:        Secret (a volume populated by a Secret)
    SecretName:  gcp-key
    Optional:    false
  default-token-s4jgd:
    Type:        Secret (a volume populated by a Secret)
    SecretName:  default-token-s4jgd
    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  67s   default-scheduler        Successfully assigned default/sv-premier-5cc8f599f6-9lrtq to docker-desktop
  Normal  Pulling    66s   kubelet, docker-desktop  Pulling image "gcr.io/proto/premiercore1:latest"
  Normal  Pulled     64s   kubelet, docker-desktop  Successfully pulled image "gcr.io/proto/premiercore1:latest"
  Normal  Created    64s   kubelet, docker-desktop  Created container sv-premier
  Normal  Started    64s   kubelet, docker-desktop  Started container sv-premier

为什么我得到这个——

Tolerations:     node.kubernetes.io/not-ready:NoExecute for 300s
                 node.kubernetes.io/unreachable:NoExecute for 300s

比我更有经验的人请帮忙

Answer 1

注意Kubernetes 会自动为node.kubernetes.io/not-ready和添加容忍度node.kubernetes.io/unreachable。

Kubernetes 会自动为 node.kubernetes.io/not-ready 添加一个 tolerationSeconds=300 的容忍度，除非用户提供的 pod 配置已经有一个对 node.kubernetes.io/not-ready 的容忍度。同样，它为 node.kubernetes.io/unreachable 添加了一个 tolerationSeconds=300 的容忍度，除非用户提供的 pod 配置已经有一个对 node.kubernetes.io/unreachable 的容忍度。

这些自动添加的容忍度确保在检测到这些问题之一后保持绑定 5 分钟的默认 pod 行为。

在这里阅读完整的细节

内置以下污点：

node.kubernetes.io/not-ready: Node is not ready. 这对应于 NodeCondition Ready 为“False”。

node.kubernetes.io/unreachable：节点无法从节点控制器访问。这对应于 NodeCondition Ready 为“未知”。

更多如下：

node.kubernetes.io/out-of-disk: 节点磁盘不足。

node.kubernetes.io/memory-pressure: 节点有内存压力。

node.kubernetes.io/disk-pressure: 节点有磁盘压力。

node.kubernetes.io/network-unavailable: 节点的网络不可用。

node.kubernetes.io/unschedulable: 节点不可调度。

node.cloudprovider.kubernetes.io/uninitialized：当 kubelet 使用“外部”云提供商启动时，此污点会在节点上设置以将其标记为不可用。在来自 cloud-controller-manager 的控制器初始化此节点后，kubelet 会删除此污点。