Spec Cost Prediction API
post
http://<your-kubecost-address>
/model/prediction/speccost
Predict API
The API requires that workloads be passed in the request body in YAML format and that the
Content-Type
header be set to application/yaml
. Multiple workloads can be passed via separation with the standard ---
syntax.Currently supported workload types:
- Deployments
- StatefulSets
- Pods
Write some Kubernetes specs to a file called
/tmp/testspecs.yaml
:read -r -d '' WL << EndOfMessage
apiVersion: apps/v1
kind: Deployment
metadata:
name: kubecost-cost-analyzer
namespace: kubecost
labels:
app: kubecost-cost-analyzer
spec:
replicas: 3
selector:
matchLabels:
app: kubecost-cost-analyzer
template:
metadata:
labels:
app: kubecost-cost-analyzer
spec:
containers:
- name: cost-model
image: nginx:1.14.2
resources:
requests:
cpu: "1m"
memory: "1Mi"
- name: cost-analyzer-frontend
image: nginx:1.14.2
resources:
requests:
cpu: "1m"
memory: "1Mi"
---
apiVersion: apps/v1
kind: Deployment
metadata:
name: default-deployment
labels:
app: default-deployment
spec:
replicas: 3
selector:
matchLabels:
app: default-deployment
template:
metadata:
labels:
app: default-deployment
spec:
containers:
- name: container-1
image: nginx:1.14.2
resources:
requests:
cpu: "10m"
memory: "10Mi"
EndOfMessage
echo "${WL}" > /tmp/testspecs.yaml
Call the endpoint with cURL, passing the file in the request body:
Request
Response
curl
-XPOST
'http://localhost:9090/model/prediction/speccost?clusterID=cluster-one&defaultNamespace=customdefault'
-H 'Content-Type: application/yaml'
--data-binary "@/tmp/testspecs.yaml"
| jq
[
{
"namespace": "kubecost",
"controllerKind": "deployment",
"controllerName": "kubecost-cost-analyzer",
"costBefore": {
"totalMonthlyRate": 3.5397661399108418,
"cpuMonthlyRate": 2.3273929838395513,
"ramMonthlyRate": 1.2123731560712905,
"gpuMonthlyRate": 0,
"monthlyCPUCoreHours": 73,
"monthlyRAMByteHours": 304653271040,
"monthlyGPUHours": 0
},
"costAfter": {
"totalMonthlyRate": 2.623504800996625,
"cpuMonthlyRate": 0.6283961056366789,
"ramMonthlyRate": 1.9951086953599462,
"gpuMonthlyRate": 0,
"monthlyCPUCoreHours": 19.71,
"monthlyRAMByteHours": 501344315550,
"monthlyGPUHours": 0
},
"costChange": {
"totalMonthlyRate": -0.9162613389142167,
"cpuMonthlyRate": -1.6989968782028724,
"ramMonthlyRate": 0.7827355392886557,
"gpuMonthlyRate": 0,
"monthlyCPUCoreHours": -53.29,
"monthlyRAMByteHours": 196691044510,
"monthlyGPUHours": 0
}
},
{
"namespace": "customdefault",
"controllerKind": "deployment",
"controllerName": "default-deployment",
"costBefore": {
"totalMonthlyRate": 0,
"cpuMonthlyRate": 0,
"ramMonthlyRate": 0,
"gpuMonthlyRate": 0,
"monthlyCPUCoreHours": 0,
"monthlyRAMByteHours": 0,
"monthlyGPUHours": 0
},
"costAfter": {
"totalMonthlyRate": 0.7896028064135204,
"cpuMonthlyRate": 0.6982178951518654,
"ramMonthlyRate": 0.09138491126165506,
"gpuMonthlyRate": 0,
"monthlyCPUCoreHours": 21.9,
"monthlyRAMByteHours": 22963814400,
"monthlyGPUHours": 0
},
"costChange": {
"totalMonthlyRate": 0.7896028064135204,
"cpuMonthlyRate": 0.6982178951518654,
"ramMonthlyRate": 0.09138491126165506,
"gpuMonthlyRate": 0,
"monthlyCPUCoreHours": 21.9,
"monthlyRAMByteHours": 22963814400,
"monthlyGPUHours": 0
}
}
]
The output will be broken down into three primary categories:
costBefore
: Represents the current monthly cost. This will be0
if the deployment is not currently running.costAfter
: The monthly cost after the change is applied.costChange
: The difference between the values ofcostBefore
andcostAfter
. If the value ofcostBefore
was0
, thencostChange
should be equal tocostAfter
.
Observe how
defaultNamespace
impacts the default-deployment
workload.From that output,
costChange
notices the existing kubecost-cost-analyzer
deployment in the kubecost
namespace and is producing an estimated negative cost difference because the request is being reduced. However, because historical usage is also factored in, there is no drastic cost reduction that might be initially expected from a 1m
CPU and 1Mi
memory request.Last modified 3mo ago