Kubecost

Product and technical docs

View project on GitHub

Kubernetes Cost Allocation

The Kubecost Allocation view allows you to quickly see allocated spend across all native Kubernetes concepts, e.g. namespace, k8s label, and service. It also allows for allocating cost to organizational concepts like team, product/project, department, or environment. This document explains the metrics presented and describes how you can control the data displayed in this view.

Cost allocation view

1. Displayed metrics

View either cumulative or run rate costs measured over the selected time window based on the resources allocated.

  • Cumulative Costs – represents the actual/historical spend captured by Kubecost agent
  • Rate metrics – hourly, daily, or monthly “run rate” cost, also used for projected cost figures, based on samples in the selected time window

Costs allocations are based on the following:

1) resources allocated, i.e. max of resource requests and usage
2) the cost of each resource
3) the amount of time resources were provisioned

For more information, refer to this FAQ on how each of these inputs is determined based on your environment.

2. Aggregation

Aggregate cost by namespace, deployment, service and other native Kubernetes concepts.

Costs aggregations are also visible by other meaningful organizational concepts, e.g. Team, Department, or Product. These aggregations are based on Kubernetes labels or annotations, referenced at both the pod and namespace-level, with labels at the pod-level being favored over the namespace label when both are present. The label name used for these concepts can be configured in Settings. Workloads without the relevent label/annotation will be shown as _unallocated_.

To find what pods are not part of the relevant label set… you can explore variations of the following kubectl commands:

kubectl get pods -l 'app notin (prometheus, cost-analyzer, ...)' --all-namespaces
kubectl get pods --show-labels -n <TARGET_NAMESPACE>

3. Time window

The designated time window for measuring costs. Results for 1d, 2d, 7d, and 30d queries are cached by default.

4. Filter

Filter resources by namespace, clusterId, and Kubernetes label to more closely investigate a rise in spend or key cost drivers at different aggregations, e.g. Deployments or Pods. When a filter is applied, only resources with this matching value will be shown. These filters are also applied to external out-of-cluster asset tags.

5. Allocate Idle Cost

Allocating idle costs proportionately distributes slack or idle cluster costs to tenants. As an example, if your cluster is only 25% utilized, as measured by the max of resource usage and requests, applying idle costs would increase the cost of each pod/namespace/deployment by 4x. This feature can be enabled by default in Settings.

6. Chart selection

Toggle to the bar chart view to see aggregated costs over the selected window, or the time series view to see cost changes over time.

7. Additional options

View other options to export cost data to CSV or view help documentation.

Cost metrics

Cost allocation metrics are available for both in-cluster and out-of-cluster resources. Here are short descriptions of each metric:

Metric Description
Memory cost The total cost of memory allocated to this object, e.g. namespace or deployment. The amount of memory allocated is the greater of memory usage and memory requested over the measured time window. The price of allocated memory is based on cloud billing APIs or custom pricing sheets. Learn more
CPU Cost The total cost of CPU allocated to this object, e.g. namespace or deployment. The amount of CPU allocated is the greater of CPU usage and CPU requested over the measured time window. The price of allocated CPU is based on cloud billing APIs or custom pricing sheets. Learn more
Cost Efficiency The percentage of requested CPU & memory dollars utilizated over the measured time window. Values range from 0 to 100 percent. Workloads without requests but with usage report 100% efficiency.
Network Cost The cost of network traffic based on internet egress, cross-zone egress, and other billed transfer. Note: these costs must be enabled. Learn more
PV Cost The cost of persistent storage volumes claimed by this object. Prices are based on cloud billing prices or custom pricing sheets for on-prem deployments.
GPU Cost The cost of GPUs requested by this object, as measured by resource limits. Prices are based on cloud billing prices or custom pricing sheets for on-prem deployments.
External Cost The cost of out-of-cluster resources allocated to this object. For example, S3 buckets allocated to a particular Kubernetes deployment. Prices are based on cloud billing data and require a key. This feature is currently available for AWS (learn more) and GCP (learn more).