Nebius Observability Agent for Kubernetes®

• Out-of-the-box metrics collection with minimal configuration
• Built-in metric enrichment (for example, adds cluster ID and other metadata)
• High reliability and tight integration with Nebius AI Cloud
• Unified observability agent that can collect logs, metrics and traces

​

Installation

​

Basic installation

1. Connect to your Managed Service for Kubernetes cluster.
2. Install the Nebius Observability Agent for Kubernetes:
   helm install nebius-observability-agent oci://cr.nebius.cloud/observability/public/nebius-observability-agent-helm \
     --version $(curl https://nebius-observability-agent.storage.eu-north1.nebius.cloud/nebius-observability-agent-helm/latest-release) \
     --namespace observability --create-namespace
   
   To use a specific agent version, change the --version value, for example:
   helm install nebius-observability-agent oci://cr.nebius.cloud/observability/public/nebius-observability-agent-helm \
     --version 1.0.4 \
     --namespace observability --create-namespace
   
3. Check the installation:
   kubectl get pods --namespace observability
   

​

Installation with custom configuration

helm install nebius-observability-agent oci://cr.nebius.cloud/observability/public/nebius-observability-agent-helm \
  --version $(curl https://nebius-observability-agent.storage.eu-north1.nebius.cloud/nebius-observability-agent-helm/latest-release) \
  --namespace observability --create-namespace \
  --values values.yaml

​

Configuration

​

Basic configuration

config:
  logs:
    enabled: true
    collectAgentLogs: false
    excludedNamespaces:
      - kube-system

  metrics:
    enabled: true
    collectAgentMetrics: false
    collectK8sClusterMetrics: false

​

Configuration parameters

​

Log collection settings

• config.logs.enabled: Enable or disable log collection. Default: true.
• config.logs.collectAgentLogs: Parameter to collect logs from the Nebius Observability Agent for Kubernetes itself. Default: false.
• config.logs.excludedNamespaces: List of namespaces to exclude from log collection.

​

Metrics collection settings

• config.metrics.enabled: Enable or disable metrics collection. Default: true.
• config.metrics.collectAgentMetrics: Parameter to collect metrics from the Nebius Observability Agent for Kubernetes itself. Default: false.
• config.metrics.collectK8sClusterMetrics: Parameter to collect Kubernetes cluster-level metrics. Default: false.
• config.metrics.excludedNamespaces: List of namespaces to exclude from metrics collection.
• config.metrics.labelAllowlist: Regular expressions that control which labels from Pod, Service and Node objects are attached to scraped metrics. Each field (podLabels, serviceLabels, nodeLabels) is matched against label names.
  Example

  config:
    metrics:
      labelAllowlist:
        podLabels: ".+"
        serviceLabels: ".+"
        nodeLabels: "kubernetes_io_hostname|topology_kubernetes_io_.*|node_kubernetes_io_instance_type|nvidia_com_gpu_present|nebius_com_.*"
  
  Default: ".+" (all labels are attached).
• config.metrics.batch.maxSizeBytes: Maximum size, in bytes, of a single metrics export request. Batches exceeding the limit are split before sending. Default: 62914560 (60 MiB).
• config.metrics.batch.maxSizeMessages: Maximum number of data points in a single metrics export request. Batches exceeding the limit are split before sending. Default: 500000.

​

Traces collection settings

• config.traces.enabled: Enable or disable trace collection. Default: true.

nebius-observability-agent.<namespace>.svc.cluster.local:4317

​

Full observability configuration

config:
  logs:
    enabled: true
    collectAgentLogs: false
    excludedNamespaces:
      - kube-system

  metrics:
    enabled: true
    collectAgentMetrics: false
    collectK8sClusterMetrics: false

  traces:
    enabled: true

​

Logs-only configuration

config:
  logs:
    enabled: true
    collectAgentLogs: false
    excludedNamespaces:
      - kube-system

  metrics:
    enabled: false

​

Metrics-only configuration

config:
  logs:
    enabled: false

  metrics:
    enabled: true
    collectAgentMetrics: false
    collectK8sClusterMetrics: false

​

Combined logs and metrics configuration

config:
  logs:
    enabled: true
    collectAgentLogs: false
    excludedNamespaces:
      - kube-system

  metrics:
    enabled: true
    collectAgentMetrics: false
    collectK8sClusterMetrics: false

​

Traces-only configuration

config:
  logs:
    enabled: false

  metrics:
    enabled: false

  traces:
    enabled: true

​

Namespace exclusions

config:
  logs:
    excludedNamespaces:
      - kube-system
      - monitoring
      - cert-manager
      - istio-system

  metrics:
    excludedNamespaces:
      - kube-system
      - monitoring
      - cert-manager
      - istio-system

​

Updating the agent

helm upgrade nebius-observability-agent oci://cr.nebius.cloud/observability/public/nebius-observability-agent-helm \
  --version $(curl https://nebius-observability-agent.storage.eu-north1.nebius.cloud/nebius-observability-agent-helm/latest-release) \
  --namespace observability \
  --values values.yaml

​

Uninstalling the agent

helm uninstall nebius-observability-agent --namespace observability

​

Troubleshooting

1. Check the agent pod status:
   kubectl get pods --namespace observability
   kubectl describe pod <Pod_name> --namespace observability
   
2. View agent logs:
   kubectl logs <Pod_name> --namespace observability
   
3. Verify the configuration:
   helm get values nebius-observability-agent --namespace observability
   

​

See also

• How to ingest metrics with the agent
• How to ingest logs with the agent
• How to send traces to the agent
• Agent changelog