为什么 pod 会自行终止？

Question

我正在尝试使用 bitnami helm chat 安装 fluend with elasticsearch 和 kibana。

我正在关注下面提到的文章

Integrate Logging Kubernetes Kibana ElasticSearch Fluentd

但是当我部署 elasticsearch 时，它的 pod 继续 Terminating 或 Back-off 状态。

我从 3 天起就一直卡在这个问题上，感谢任何帮助。

事件：

  Type     Reason            Age                From               Message
  ----     ------            ----               ----               -------
  Warning  FailedScheduling  41m (x2 over 41m)  default-scheduler  error while running "VolumeBinding" filter plugin for pod "elasticsearch-master-0": pod has unbound immediate PersistentVolumeClaims
  Normal   Scheduled         41m                default-scheduler  Successfully assigned default/elasticsearch-master-0 to minikube
  Normal   Pulling           41m                kubelet, minikube  Pulling image "busybox:latest"
  Normal   Pulled            41m                kubelet, minikube  Successfully pulled image "busybox:latest"
  Normal   Created           41m                kubelet, minikube  Created container sysctl
  Normal   Started           41m                kubelet, minikube  Started container sysctl
  Normal   Pulling           41m                kubelet, minikube  Pulling image "docker.elastic.co/elasticsearch/elasticsearch-oss:6.8.6"
  Normal   Pulled            39m                kubelet, minikube  Successfully pulled image "docker.elastic.co/elasticsearch/elasticsearch-oss:6.8.6"
  Normal   Created           39m                kubelet, minikube  Created container chown
  Normal   Started           39m                kubelet, minikube  Started container chown
  Normal   Created           38m                kubelet, minikube  Created container elasticsearch
  Normal   Started           38m                kubelet, minikube  Started container elasticsearch
  Warning  Unhealthy         38m                kubelet, minikube  Readiness probe failed: Get http://172.17.0.7:9200/_cluster/health?local=true: dial tcp 172.17.0.7:9200: connect: connection refused
  Normal   Pulled            38m (x2 over 38m)  kubelet, minikube  Container image "docker.elastic.co/elasticsearch/elasticsearch-oss:6.8.6" already present on machine
  Warning  FailedMount       32m                kubelet, minikube  MountVolume.SetUp failed for volume "config" : failed to sync configmap cache: timed out waiting for the condition
  Normal   SandboxChanged    32m                kubelet, minikube  Pod sandbox changed, it will be killed and re-created.
  Normal   Pulling           32m                kubelet, minikube  Pulling image "busybox:latest"
  Normal   Pulled            32m                kubelet, minikube  Successfully pulled image "busybox:latest"
  Normal   Created           32m                kubelet, minikube  Created container sysctl
  Normal   Started           32m                kubelet, minikube  Started container sysctl
  Normal   Pulled            32m                kubelet, minikube  Container image "docker.elastic.co/elasticsearch/elasticsearch-oss:6.8.6" already present on machine
  Normal   Created           32m                kubelet, minikube  Created container chown
  Normal   Started           32m                kubelet, minikube  Started container chown
  Normal   Pulled            32m (x2 over 32m)  kubelet, minikube  Container image "docker.elastic.co/elasticsearch/elasticsearch-oss:6.8.6" already present on machine
  Normal   Created           32m (x2 over 32m)  kubelet, minikube  Created container elasticsearch
  Normal   Started           32m (x2 over 32m)  kubelet, minikube  Started container elasticsearch
  Warning  Unhealthy         32m                kubelet, minikube  Readiness probe failed: Get http://172.17.0.6:9200/_cluster/health?local=true: dial tcp 172.17.0.6:9200: connect: connection refused
  Warning  BackOff           32m (x2 over 32m)  kubelet, minikube  Back-off restarting failed container

Answer 1

简答：它崩溃了。您可以检查 Pod 状态对象以获取一些详细信息，例如退出状态以及是否是 oomkill，然后查看容器日志以查看它们是否显示任何内容。

Answer 2

这里的问题是 pod 具有未绑定的即时 PersistentVolumeClaims。您可以在使用 helm 部署它时将 master.persistence.enabled 设置为 false。或者，您需要检查集群中是否存在默认存储 class，如果不存在，则创建存储 class 并将其设为默认。

为什么 pod 会自行终止？

Why pod terminate it self?

azure

fluentd

kubernetes

kubernetes-helm