How to Enable cAdvisor Metrics in Observe Agent

Data Ingestion
, , ,
1

:brain: Enabling cAdvisor Metrics for Your Node Monitoring

If you’re looking to monitor node- and container-level performance metrics such as CPU, memory, filesystem, and network usage, cAdvisor is your go-to. It’s built into the Kubernetes kubelet and can easily be enabled via your Observe Agent configuration.

:gear: Step 1: Enable cAdvisor Metrics

Create a values.yaml (used by your Helm deployment of the Observe Agent), add the following section:

node:
  metrics:
    cadvisor:
      enabled: true

This configuration tells the Observe Agent to collect metrics from the cAdvisor endpoint on each node.

:rocket: Step 2: Upgrade the Observe Agent

Once you’ve confirmed your configuration, apply it by upgrading your Helm release:

helm upgrade observe-agent observe/agent -n observe -f values.yaml

This command:

  • Uses your updated values.yaml to apply configuration changes.
  • Upgrades (or redeploys) the Observe Agent in the observe namespace.
  • Ensures the new settings, including cAdvisor metrics collection, are applied.

To confirm the upgrade worked, you can check the pod status:

kubectl get pods -n observe

And verify the running configuration:

helm get values observe-agent -n observe

:puzzle_piece: Step 3: Validate Metrics Collection

Once the upgrade is complete:

  1. Go to your Observe dashboard or metrics visualization tool.
  2. Check for new cAdvisor metrics under node or container scopes — such as:
  • container_cpu_usage_seconds_total
  • container_memory_usage_bytes
  • container_fs_usage_bytes

If you see these, you’re successfully collecting cAdvisor metrics :tada:.