Kubernetes distributions

Overview

This section aims to document specifics and to provide good base configuration for all major Kubernetes distributions. These configuration can then be customized to add any Datadog feature.

AWS Elastic Kubernetes Service (EKS)
Azure Kubernetes Service (AKS)
Google Kubernetes Engine (GKE)
Red Hat OpenShift
Rancher
Oracle Container Engine for Kubernetes (OKE)
vSphere Tanzu Kubernetes Grid (TKG)

AWS Elastic Kubernetes Service (EKS)

No specific configuration is required.

If you are using AWS Bottlerocket OS on your nodes, add the following to enable container monitoring (containerd check):

Custom values.yaml:

datadog:
  apiKey: <DATADOG_API_KEY>
  appKey: <DATADOG_APP_KEY>
  criSocketPath: /run/dockershim.sock
  env:
  - name: DD_AUTOCONFIG_INCLUDE_FEATURES
    value: "containerd"

DatadogAgent Kubernetes Resource:

apiVersion: datadoghq.com/v1alpha1
kind: DatadogAgent
metadata:
  name: datadog
spec:
  credentials:
    apiKey: <DATADOG_API_KEY>
    appKey: <DATADOG_APP_KEY>
  agent:
    config:
      criSocket:
        criSocketPath: /run/dockershim.sock
  clusterAgent:
    image:
      name: "gcr.io/datadoghq/cluster-agent:latest"
    config:
      externalMetrics:
        enabled: false
      admissionController:
        enabled: false

Azure Kubernetes Service (AKS)

AKS requires specific configuration for the Kubelet integration due to AKS certificates setup.

Custom values.yaml:

datadog:
  apiKey: <DATADOG_API_KEY>
  appKey: <DATADOG_APP_KEY>
  kubelet:
    tlsVerify: false # Required as of Agent 7.35. See Notes.

DatadogAgent Kubernetes Resource:

apiVersion: datadoghq.com/v1alpha1
kind: DatadogAgent
metadata:
  name: datadog
spec:
  credentials:
    apiKey: <DATADOG_API_KEY>
    appKey: <DATADOG_APP_KEY>
  agent:
    config:
      kubelet:
        tlsVerify: false # Required as of Agent 7.35. See Notes.
  clusterAgent:
    image:
      name: "gcr.io/datadoghq/cluster-agent:latest"
    config:
      externalMetrics:
        enabled: false
      admissionController:
        enabled: false

Notes:

As of Agent 7.35, tlsVerify: false is required because Kubelet certificates in AKS do not have a Subject Alternative Name (SAN) set.
In some setups, DNS resolution for spec.nodeName inside Pods may not work in AKS. This has been reported on all AKS Windows nodes and when cluster is setup in a Virtual Network using custom DNS on Linux nodes. In this case, removing the agent.config.kubelet.host field (defaults to status.hostIP) and using tlsVerify: false is required. Using the DD_KUBELET_TLS_VERIFY=false environment variable also resolves this issue. Both of these options deactivate verification of the server certificate.
```
env:
  - name: DD_KUBELET_TLS_VERIFY
    value: "false"
```
Admission Controller functionality on AKS requires configuring the add selectors to prevent an error on reconciling the webhook:

clusterAgent:
  env:
    - name: "DD_ADMISSION_CONTROLLER_ADD_AKS_SELECTORS"
      value: "true"

Google Kubernetes Engine (GKE)

GKE can be configured in two different mode of operation:

Standard: You manage the cluster’s underlying infrastructure, giving you node configuration flexibility.
Autopilot: GKE provisions and manages the cluster’s underlying infrastructure, including nodes and node pools, giving you an optimized cluster with a hands-off experience.

Depending on the operation mode of your cluster, the Datadog Agent needs to be configured differently.

Standard

Since Agent 7.26, no specific configuration is required for GKE (whether you run Docker or containerd).

Note: When using COS (Container Optimized OS), the eBPF-based OOM Kill and TCP Queue Length checks are supported starting from the version 3.0.1 of the Helm chart. To enable these checks, configure the following setting:

datadog.systemProbe.enableDefaultKernelHeadersPaths to false.

Autopilot

GKE Autopilot requires some configuration, shown below.

Datadog recommends that you specify resource limits for the Agent container. Autopilot sets a relatively low default limit (50m CPU, 100Mi memory) that may quickly lead the Agent container to OOMKill depending on your environment. If applicable, also specify resource limits for the Trace Agent and Process Agent containers.

Custom values.yaml:

datadog:
  apiKey: <DATADOG_API_KEY>
  appKey: <DATADOG_APP_KEY>
  clusterName: <CLUSTER_NAME>

  # Enable the new `kubernetes_state_core` check.
  kubeStateMetricsCore:
    enabled: true
  # Avoid deploying kube-state-metrics chart.
  # The new `kubernetes_state_core` doesn't require to deploy the kube-state-metrics anymore.
  kubeStateMetricsEnabled: false

agents:
  containers:
    agent:
      # resources for the Agent container
      resources:
        requests:
          cpu: 200m
          memory: 256Mi
        limits:
          cpu: 200m
          memory: 256Mi

    traceAgent:
      # resources for the Trace Agent container
      resources:
        requests:
          cpu: 100m
          memory: 200Mi
        limits:
          cpu: 100m
          memory: 200Mi

    processAgent:
      # resources for the Process Agent container
      resources:
        requests:
          cpu: 100m
          memory: 200Mi
        limits:
          cpu: 100m
          memory: 200Mi

providers:
  gke:
    autopilot: true

Red Hat OpenShift

OpenShift comes with hardened security by default (SELinux, SecurityContextConstraints), thus requiring some specific configuration:

Create SCC for Node Agent and Cluster Agent
Specific CRI socket path as OpenShift uses CRI-O container runtime
Kubelet API certificates may not always be signed by cluster CA
Tolerations are required to schedule the Node Agent on master and infra nodes
Cluster name should be set as it cannot be retrieved automatically from cloud provider

This configuration supports OpenShift 3.11 and OpenShift 4, but works best with OpenShift 4.

Custom values.yaml:

datadog:
  apiKey: <DATADOG_API_KEY>
  appKey: <DATADOG_APP_KEY>
  clusterName: <CLUSTER_NAME>
  criSocketPath: /var/run/crio/crio.sock
  # Depending on your DNS/SSL setup, it might not be possible to verify the Kubelet cert properly
  # If you have proper CA, you can switch it to true
  kubelet:
    tlsVerify: false
agents:
  podSecurity:
    securityContextConstraints:
      create: true
  tolerations:
  - effect: NoSchedule
    key: node-role.kubernetes.io/master
    operator: Exists
  - effect: NoSchedule
    key: node-role.kubernetes.io/infra
    operator: Exists
clusterAgent:
  podSecurity:
    securityContextConstraints:
      create: true
kube-state-metrics:
  securityContext:
    enabled: false

When using the Datadog Operator in OpenShift, it is recommended that you install it through OperatorHub or RedHat Marketplace. The configuration below is meant to work with this setup (due to SCC/ServiceAccount setup), when the Agent is installed in the same namespace as the Datadog Operator.

apiVersion: datadoghq.com/v1alpha1
kind: DatadogAgent
metadata:
  name: datadog
spec:
  credentials:
    apiKey: <DATADOG_API_KEY>
    appKey: <DATADOG_APP_KEY>
  clusterName: <CLUSTER_NAME>
  agent:
    image:
      name: "gcr.io/datadoghq/agent:latest"
    apm:
      enabled: false
    process:
      enabled: true
      processCollectionEnabled: false
    log:
      enabled: false
    systemProbe:
      enabled: false
    security:
      compliance:
        enabled: false
      runtime:
        enabled: false
    rbac:
      serviceAccountName: datadog-agent-scc
    config:
      kubelet:
        tlsVerify: false
      criSocket:
        criSocketPath: /var/run/crio/crio.sock
        useCriSocketVolume: true
      tolerations:
      - effect: NoSchedule
        key: node-role.kubernetes.io/master
        operator: Exists
      - effect: NoSchedule
        key: node-role.kubernetes.io/infra
        operator: Exists
  clusterAgent:
    image:
      name: "gcr.io/datadoghq/cluster-agent:latest"
    config:
      externalMetrics:
        enabled: false
        port: 8443
      admissionController:
        enabled: false

Rancher

Rancher installations are close to vanilla Kubernetes, requiring only some minor configuration:

Tolerations are required to schedule the Node Agent on controlplane and etcd nodes
Cluster name should be set as it cannot be retrieved automatically from cloud provider

Custom values.yaml:

datadog:
  apiKey: <DATADOG_API_KEY>
  appKey: <DATADOG_APP_KEY>
  clusterName: <CLUSTER_NAME>
  kubelet:
    tlsVerify: false
agents:
  tolerations:
  - effect: NoSchedule
    key: node-role.kubernetes.io/controlplane
    operator: Exists
  - effect: NoExecute
    key: node-role.kubernetes.io/etcd
    operator: Exists

DatadogAgent Kubernetes Resource:

apiVersion: datadoghq.com/v1alpha1
kind: DatadogAgent
metadata:
  name: datadog
spec:
  credentials:
    apiKey: <DATADOG_API_KEY>
    appKey: <DATADOG_APP_KEY>
  clusterName: <CLUSTER_NAME>
  agent:
    image:
      name: "gcr.io/datadoghq/agent:latest"
    apm:
      enabled: false
    process:
      enabled: true
      processCollectionEnabled: false
    log:
      enabled: false
    systemProbe:
      enabled: false
    security:
      compliance:
        enabled: false
      runtime:
        enabled: false
    config:
      kubelet:
        tlsVerify: false
      tolerations:
      - effect: NoSchedule
        key: node-role.kubernetes.io/controlplane
        operator: Exists
      - effect: NoExecute
        key: node-role.kubernetes.io/etcd
        operator: Exists
  clusterAgent:
    image:
      name: "gcr.io/datadoghq/cluster-agent:latest"
    config:
      externalMetrics:
        enabled: false
      admissionController:
        enabled: false

Oracle Container Engine for Kubernetes (OKE)

No specific configuration is required.

To enable container monitoring, add the following (containerd check):

Custom values.yaml:

datadog:
  apiKey: <DATADOG_API_KEY>
  appKey: <DATADOG_APP_KEY>
  criSocketPath: /run/dockershim.sock
  env:
  - name: DD_AUTOCONFIG_INCLUDE_FEATURES
    value: "containerd"

DatadogAgent Kubernetes Resource:

apiVersion: datadoghq.com/v1alpha1
kind: DatadogAgent
metadata:
  name: datadog
spec:
  credentials:
    apiKey: <DATADOG_API_KEY>
    appKey: <DATADOG_APP_KEY>
  agent:
    config:
      criSocket:
        criSocketPath: /run/dockershim.sock
  clusterAgent:
    image:
      name: "gcr.io/datadoghq/cluster-agent:latest"
    config:
      externalMetrics:
        enabled: false
      admissionController:
        enabled: false

More values.yaml examples can be found in the Helm chart repository More DatadogAgent examples can be found in the Datadog Operator repository

vSphere Tanzu Kubernetes Grid (TKG)

TKG requires some small configuration changes, shown below. For example, setting a toleration is required for the controller to schedule the Node Agent on the master nodes.

Custom values.yaml:

datadog:
  apiKey: <DATADOG_API_KEY>
  appKey: <DATADOG_APP_KEY>
  kubelet:
    # Set tlsVerify to false since the Kubelet certificates are self-signed
    tlsVerify: false
  # Disable the `kube-state-metrics` dependency chart installation.
  kubeStateMetricsEnabled: false
  # Enable the new `kubernetes_state_core` check.
  kubeStateMetricsCore:
    enabled: true
# Add a toleration so that the agent can be scheduled on the control plane nodes.
agents:
  tolerations:
    - key: node-role.kubernetes.io/master
      effect: NoSchedule

DatadogAgent Kubernetes Resource:

apiVersion: datadoghq.com/v1alpha1
kind: DatadogAgent
metadata:
  name: datadog
spec:
  credentials:
    apiSecret:
      secretName: datadog-secret
      keyName: api-key
    appSecret:
      secretName: datadog-secret
      keyName: app-key
  features:
    # Enable the new `kubernetes_state_core` check.
    kubeStateMetricsCore:
      enabled: true
  agent:
    config:
      kubelet:
        # Set tlsVerify to false since the Kubelet certificates are self-signed
        tlsVerify: false
      # Add a toleration so that the agent can be scheduled on the control plane nodes.
      tolerations:
        - key: node-role.kubernetes.io/master
          effect: NoSchedule
  clusterAgent:
    config:
      collectEvents: true