Kubernetes

Supported OS Linux Mac OS

통합 버전1.7.0
이 페이지는 아직 영어로 제공되지 않습니다. 번역 작업 중입니다.
현재 번역 프로젝트에 대한 질문이나 피드백이 있으신 경우 언제든지 연락주시기 바랍니다.

Kubernetes Dashboard

Overview

Get metrics from the Kubernetes service in real time to:

  • Visualize and monitor Kubernetes states
  • Be notified about Kubernetes failovers and events.

Note: This check only works with Agent v5. For Agent v6+, see the kubelet check.

Setup

Installation

The Kubernetes check is included in the Datadog Agent package, so you don’t need to install anything else on your Kubernetes servers.

For more information on installing the Datadog Agent on your Kubernetes clusters, see the Kubernetes documentation.

To collect Kubernetes State metrics, see the kubernetes_state integration.

Configuration

Edit the kubernetes.yaml file to point to your server and port, set the masters to monitor.

Validation

Run the Agent’s status subcommand and look for kubernetes under the Checks section.

Data Collected

Metrics

kubernetes.cpu.capacity
(gauge)		The number of cores in this machine
Shown as core
kubernetes.cpu.limits
(gauge)		The limit of cpu cores set
Shown as core
kubernetes.cpu.requests
(gauge)		The requested cpu cores
Shown as core
kubernetes.cpu.usage.total
(gauge)		The number of cores used
Shown as nanocore
kubernetes.diskio.io_service_bytes.stats.total
(gauge)		The amount of disk space the container uses.
Shown as byte
kubernetes.filesystem.usage
(gauge)		The amount of disk used. Requires Docker container runtime.
Shown as byte
kubernetes.filesystem.usage_pct
(gauge)		The percentage of disk used. Requires Docker container runtime.
Shown as fraction
kubernetes.memory.capacity
(gauge)		The amount of memory (in bytes) in this machine
Shown as byte
kubernetes.memory.limits
(gauge)		The limit of memory set
Shown as byte
kubernetes.memory.requests
(gauge)		The requested memory
Shown as byte
kubernetes.memory.usage
(gauge)		The amount of memory used
Shown as byte
kubernetes.network.rx_bytes
(gauge)		The amount of bytes per second received
Shown as byte
kubernetes.network.tx_bytes
(gauge)		The amount of bytes per second transmitted
Shown as byte
kubernetes.network_errors
(gauge)		The amount of network errors per second
Shown as error

Events

As of the v5.17.0 release, the Datadog Agent supports a built-in leader election option for the Kubernetes event collector. Once enabled, you no longer need to deploy an additional event collection container to your cluster. Instead, Agents coordinate to ensure only one Agent instance is gathering events at a given time, events below are available:

  • Backoff
  • Conflict
  • Delete
  • DeletingAllPods
  • Didn’t have enough resource
  • Error
  • Failed
  • FailedCreate
  • FailedDelete
  • FailedMount
  • FailedSync
  • Failedvalidation
  • FreeDiskSpaceFailed
  • HostPortConflict
  • InsufficientFreeCPU
  • InsufficientFreeMemory
  • InvalidDiskCapacity
  • Killing
  • KubeletsetupFailed
  • NodeNotReady
  • NodeoutofDisk
  • OutofDisk
  • Rebooted
  • TerminatedAllPods
  • Unable
  • Unhealthy

Service Checks

kubernetes_state.node.ready

Returns CRITICAL if a cluster node is not ready. Returns WARNING if status is unknown. Returns OK otherwise.

Statuses: ok, warning, critical

kubernetes_state.node.out_of_disk

Returns CRITICAL if a cluster node is out of disk space. Returns UNKNOWN if status is unknown. Returns OK otherwise.

Statuses: ok, unknown, critical

kubernetes_state.node.disk_pressure

Returns CRITICAL if a cluster node is in a disk pressure state. Returns UNKNOWN if status is unknown. Returns OK otherwise.

Statuses: ok, unknown, critical

kubernetes_state.node.memory_pressure

Returns CRITICAL if a cluster node is in a memory pressure state. Returns UNKNOWN if status is unknown. Returns OK otherwise.

Statuses: ok, unknown, critical

kubernetes_state.node.network_unavailable

Returns CRITICAL if a cluster node is in a network unavailable state. Returns UNKNOWN if status is unknown. Returns OK otherwise.

Statuses: ok, unknown, critical

kubernetes_state.cronjob.on_schedule_check

Returns CRITICAL if a cron job scheduled time is unknown or in the past. Returns OK otherwise.

Statuses: ok, critical

kubernetes_state.job.complete

Returns CRITICAL if a job failed. Returns OK otherwise.

Statuses: ok, critical

kubernetes_state.cronjob.complete

Returns CRITICAL if the last job of a cronjob failed. Returns OK otherwise.

Statuses: ok, critical

Troubleshooting

Agent installation on Kubernetes master nodes

Since Kubernetes v1.6, the concept of Taints and tolerations was introduced. The master node is no longer off limits, it’s simply tainted. Add the required toleration to the pod to run it.

Add the following lines to your Deployment (or Daemonset if you are running a multi-master setup):

spec:
  tolerations:
    - key: node-role.kubernetes.io/master
      effect: NoSchedule

Why is the Kubernetes check failing with a ConnectTimeout error to port 10250?

The Agent assumes the kubelet API is available at the default gateway of the container. If that’s not the case because you are using a software defined networks like Calico or Flannel, the Agent needs to be specified using an environment variable:

- name: KUBERNETES_KUBELET_HOST
  valueFrom:
    fieldRef:
      fieldPath: spec.nodeName

For reference, see this pull request.

Why is there a container in each Kubernetes pod with 0% CPU and minimal disk/ram?

These are pause containers (docker_image:gcr.io/google_containers/pause.*) that K8s injects into every pod to keep it populated even if the “real” container is restarting or stopped.

The docker_daemon check ignores them through a default exclusion list, but they do show up for K8s metrics like kubernetes.cpu.usage.total and kubernetes.filesystem.usage.

Further Reading