Datadog Kubernetes FAQs

Does integrating Datadog Kubernetes also create Datadog as a cost center?

Yes. Datadog Kubernetes integration will also create a Datadog cost center in Finout, enabling visibility into your Datadog spend.

Are there differences in how Kubernetes cost enrichment works across AWS, GCP, and Azure?

All three cloud providers support Finout’s new CPU, memory and network metrics for calculating and breaking down Kubernetes costs across Finout.

Kubernetes costs using AWS infrastructure can be calculated by the old and the refined algorithm, based on your account configuration.

Kubernetes costs using GCP or Azure infrastructure can be calculated by the refined algorithm only.

Why are some resources not identified as Kubernetes in Finout?

Finout identifies Kubernetes resources using metrics collected from your configured monitoring source. If the required metrics for a specific node or machine are not available, Finout cannot associate that infrastructure with Kubernetes.

In such cases, the cost will still appear in Finout, but it will be attributed to the underlying cloud provider (for example, AWS, GCP, or Azure) rather than being enriched with Kubernetes dimensions such as namespace, workload, or labels.

Why don't my Kubernetes labels from Datadog not show up in Finout?

Datadog collects Kubernetes labels, but Finout can only read them if they are used as tags in Datadog. By default, labels aren’t exposed via the Datadog API. To fix this, configure your Datadog Agent/Cluster Agent to map pod, namespace, or node labels as tags. Once mapped, they’ll appear in Datadog as tags and then surface automatically in Finout under Kubernetes pod labels.

See Datadog’s guide for using Kubernetes labels as tags for setup instructions.

Why don’t some AKS nodes match Azure billing resources in Finout?

Datadog host/provider identifiers for AKS can appear in multiple formats (for example, aks-... hostnames and VMSS-style suffixes). Finout normalizes these variants, including Azure VMSS instance suffixes encoded in Base36, to improve matching between Kubernetes metrics and Azure cost data. If matches are still missing, share sample host tags/provider IDs with Finout support so we can verify the mapping strategy for your account.

How is a machine identified as Kubernetes in Finout?

For a machine to be identified as Kubernetes, specific Kubernetes-related metrics must be present. If these metrics are missing, Finout cannot classify the resource as Kubernetes and will instead attribute it to its original cloud provider.