Cluster Agent commands

The available commands for the Datadog Cluster Agents are:

datadog-cluster-agent status

Gives an overview of the components of the Agent and their health.

datadog-cluster-agent metamap <NODE_NAME>

Queries the local cache of the mapping between the pods living on NODE_NAME, and the cluster level metadata they are associated with, such as endpoints. Not specifying the NODE_NAME runs the mapper on all the nodes of the cluster.

datadog-cluster-agent flare <CASE_ID>

Similarly to the node-based Agent, the Cluster Agent can aggregate the logs and the configurations used and forward an archive to the support team, or be deflated and used locally. Note: this command runs from within the Cluster Agent pod.

Cluster Agent environment variables

apiVersion: datadoghq.com/v2alpha1
kind: DatadogAgent
metadata:
  name: datadog
spec:
  override:
    clusterAgent:
      env:
        - name: <ENV_VAR_NAME>
          value: <ENV_VAR_VALUE>

clusterAgent:
  env:
    - name: <ENV_VAR_NAME>
      value: <ENV_VAR_VALUE>

The following environment variables are supported:

DD_API_KEY

Your Datadog API key.

DD_CLUSTER_CHECKS_ENABLED

Enable Cluster Check Autodiscovery. Defaults to false.

DD_CLUSTER_AGENT_AUTH_TOKEN

32-character token that needs to be shared between the node Agent and the Datadog Cluster Agent.

DD_CLUSTER_AGENT_KUBERNETES_SERVICE_NAME

Name of the Kubernetes service through which Cluster Agents are exposed. Defaults to datadog-cluster-agent.

DD_CLUSTER_NAME

Cluster name. Added as an instance tag to all cluster check configurations.

DD_CLUSTER_CHECKS_ENABLED

When true, enables dispatching logic on the leader Cluster Agent. Default: false.

DD_CLUSTER_CHECKS_NODE_EXPIRATION_TIMEOUT

Time (in seconds) after which node-based Agents are considered down and removed from the pool. Defaults to 30 seconds.

DD_CLUSTER_CHECKS_WARMUP_DURATION

Delay (in seconds) between acquiring leadership and starting the Cluster Checks logic, allowing for all node-based Agents to register first. Default is 30 seconds.

DD_CLUSTER_CHECKS_CLUSTER_TAG_NAME

Name of the instance tag set with the DD_CLUSTER_NAME option. Defaults to cluster_name.

DD_CLUSTER_CHECKS_EXTRA_TAGS

Adds extra tags to cluster check metrics.

DD_CLUSTER_CHECKS_ADVANCED_DISPATCHING_ENABLED

When true, the leader Cluster Agent collects stats from the cluster-level check runners to optimize check dispatching logic. Default: true for Agent v7.73 or later (false for earlier versions).

DD_CLUSTER_CHECKS_KSM_SHARDING_ENABLED

Enables Kubernetes State Metrics Resource Type Sharding that automatically splits the kubernetes_state_core cluster check into multiple shards based on resource type groups, enabling parallel execution across multiple Cluster Check Runners (CLC runners) for improved performance and scalability in large Kubernetes clusters. Requires Agent v7.74 or higher and DD_CLUSTER_CHECKS_ADVANCED_DISPATCHING_ENABLED to be set to true.

DD_CLUSTER_CHECKS_CLC_RUNNERS_PORT

The port used by the Cluster Agent client to reach cluster-level check runners and collect their stats. Default: 5005.

DD_HOSTNAME

Hostname to use for the Datadog Cluster Agent.

DD_ENV

Sets the env tag for data emitted by the Cluster Agent. Recommended only if the Cluster Agent monitors services within a single environment.

DD_USE_METADATA_MAPPER

Enables cluster level metadata mapping. Defaults to true.

DD_COLLECT_KUBERNETES_EVENTS

Configures the Agent to collect Kubernetes events. Defaults to false.

DD_LEADER_ELECTION

Activates leader election. Set DD_COLLECT_KUBERNETES_EVENTS to true to activate this feature. Defaults to false.

DD_LEADER_LEASE_DURATION

Used only if leader election is activated. Value in seconds, 60 by default.

DD_KUBE_RESOURCES_NAMESPACE

Configures the namespace where the Cluster Agent creates the configmaps required for the leader election, event collection (optional), and horizontal pod autoscaling.

DD_KUBERNETES_INFORMERS_RESYNC_PERIOD

Frequency (in seconds) for querying the API server to resync the local cache. The default is 5 minutes, or 300 seconds.

DD_KUBERNETES_INFORMERS_RESTCLIENT_TIMEOUT

Timeout (in seconds) of the client communicating with the API server. Defaults to 60 seconds.

DD_METRICS_PORT

Port to expose Datadog Cluster Agent metrics. Defaults to port 5000.

DD_EXTERNAL_METRICS_PROVIDER_BATCH_WINDOW

Time waited (in seconds) to process a batch of metrics from multiple autoscalers. Defaults to 10 seconds.

DD_EXTERNAL_METRICS_PROVIDER_MAX_AGE

Maximum age (in seconds) of a datapoint before considering it invalid to be served. Default to 120 seconds.

DD_EXTERNAL_METRICS_AGGREGATOR

Aggregator for Datadog metrics. Applies to all autoscalers processed. Choose from sum/avg/max/min.

DD_EXTERNAL_METRICS_PROVIDER_BUCKET_SIZE

Size of the window (in seconds) used to query metrics from Datadog. Defaults to 300 seconds.

DD_EXTERNAL_METRICS_LOCAL_COPY_REFRESH_RATE

Rate to resync local cache of processed metrics with the global store. Useful when there are several replicas of the Cluster Agent.

DD_EXTRA_CONFIG_PROVIDERS

Additional Autodiscovery configuration providers to use.

DD_EXTRA_LISTENERS

Additional Autodiscovery listeners to run.

DD_PROXY_HTTPS

Sets a proxy server for HTTPS requests.

DD_PROXY_HTTP

Sets a proxy server for HTTP requests.

DD_PROXY_NO_PROXY

Sets a list of hosts that should bypass the proxy. The list is space-separated.

DD_ADMISSION_CONTROLLER_AUTO_INSTRUMENTATION_INIT_RESOURCES_CPU

Configures the CPU request and limit for the init containers.

DD_ADMISSION_CONTROLLER_AUTO_INSTRUMENTATION_INIT_RESOURCES_MEMORY

Configures the memory request and limit for the init containers.

Further Reading

Additional helpful documentation, links, and articles:

Introducing the Datadog Cluster AgentBLOG Autoscale your Kubernetes workloads with any Datadog metricBLOG Running Cluster Checks with AutodiscoveryDOCUMENTATION Kubernetes DaemonSet SetupDOCUMENTATION Troubleshooting the Datadog Cluster AgentDOCUMENTATION