Skip to content

allocation.md

Latest commit

 

History

History
53 lines (45 loc) · 3.82 KB

allocation.md

File metadata and controls

53 lines (45 loc) · 3.82 KB
Raw

Cost Allocation API

The cost allocation API provides workload allocation data from the daily Kubecost ETL pipeline and powers the Kubecost Reports view. Workloads can be aggregated by any Kubernetes concept, e.g. namespace, label, controller, service, pod, etc. The endpoint is available at the following address:

http://<kubecost-address>/model/allocation

Here are example uses:

API parameters include the following:

  • window dictates the applicable window for measuring historical cost. Given this API uses the Kubecost ETL pipeline, data is returned with daily resolution. For more granular cost metrics, view the aggregatedCostModel API. Supported time window options are as follows:
    • "15m", "24h", "7d", "48h", etc.
    • "today", "yesterday", "week", "month", "lastweek", "lastmonth"
    • "1586822400,1586908800", etc. (start and end unix timestamps)
    • "2020-04-01T00:00:00Z,2020-04-03T00:00:00Z", etc. (start and end UTC RFC3339 pairs)
  • aggregate is used to consolidate cost model data. Supported types are cluster, namespace, deployment, controller, service, and label. With labels, a CSV with label:key format is used.
  • accumulate when set to false this endpoint returns daily time series data vs cumulative data. Default value is false.
  • shareIdle (optional) when set to true applies the cost of all idle compute resources to tenants, default false.
  • shareCost (optional) a fixed external monthly amount to be split between tenants, e.g. 1000
  • shareNamespaces(optional) a CSV list of Kubernetes namespaces whose costs should be shared, e.g. kubecost
  • shareLabels(optional) a CSV list of Kubernetes labels whose costs should be shared, e.g. app:prometheus
  • shareSplit to be used with shareCost, shareNamespaces, or shareLabels. Supported options: weighted (to share proportionate to tenant cost) and even (to share cost uniformly)

This API returns a set of JSON objects in this format:

{
  cpuCoreHours: 0.05                     // cumulative hours of CPU cores consumed
  cpuCost: 0.083                         // total cost of CPU allocated
  cpuEfficiency: .87                     // percentage of CPU requested that is utilized, weighted by cost
  end: "2020-09-14T00:00:00+01:00"       // end of window
  gpuCost: 0                             // total cost of GPU allocated 
  gpuHours: 0                            // cumulative hours of GPU consumed, in # of GPUs
  minutes: 1440                          // cumulative count of minutes running
  name: "data-science"                   // value of aggregator 
  networkCost: 0                         // measured cost of network egress
  properties: {cluster: "cluster-one"}   // meta-data for this aggregation
  pvByteHours: 100                       // cumulative hours of disk consumed, in bytes
  pvCost: .70                            // total cost of persistent volume allocated
  ramByteHours: 123684420.26             // cumulative hours of RAM consumed, in bytes
  ramCost: 0.023                         // total cost of RAM allocated
  ramEfficiency: .56                     // percentage of RAM requested that is utilized, weighted by cost
  sharedCost: 0
  start: "2020-09-13T00:00:00+01:00"     // beginning of window
  totalCost: 0.024                       // sum of all costs
  totalEfficiency: .72                   // percentage of RAM + CPU equested that is utilized, weighted by cost
}