Skip to content

RAGEngine instanceType validation #1479

@pauldotyu

Description

@pauldotyu

Describe the bug
When attempting to deploy a new RAGEngine custom resource using the following SKU standard_d2s_v6, I see an admission webhook validation error with the following message:

Error from server (BadRequest): error when creating "STDIN": admission webhook "validation.ragengine.kaito.sh" denied the request: validation failed: invalid value: Unsupported instance type standard_d2s_v6. Supported SKUs: 

However, when redeploying with the following SKU Standard_D2S_v6 the deployment succeeds. Only change here was the capitalization of a few characters.

Steps To Reproduce

Deploy gpu-provisioner and ragengine operators into AKS cluster then attempt to deploy RAGEngine custom resource.

kubectl apply -f - <<EOF
apiVersion: kaito.sh/v1alpha1 
kind: RAGEngine 
metadata: 
  name: ragengine-llm-d 
spec: 
  compute: 
    instanceType: "standard_d2s_v6" 
    labelSelector: 
      matchLabels: 
        node.kubernetes.io/instance-type: standard_d2s_v6
  embedding:
    local:
      modelID: "BAAI/bge-small-en-v1.5"
  inferenceService:
    url: "http://$INFERENCE_SERVICE_IP/v1/chat/completions"
EOF

Expected behavior

standard_d2s_v6 and Standard_D2S_v6 should be equivalent and the error messaging should be updated since Standard_D2S_v6 deployments do work.

Logs

Error from server (BadRequest): error when creating "STDIN": admission webhook "validation.ragengine.kaito.sh" denied the request: validation failed: invalid value: Unsupported instance type standard_d2s_v6. Supported SKUs: [Standard_NC24s_v3 Standard_NC24ads_A100_v4 Standard_NCC40ads_H100_v5 Standard_NC6s_v3 Standard_ND96asr_A100_v4 Standard_NC80adis_H100_v5 Standard_NC64as_T4_v3 Standard_NC48ads_A100_v4 Standard_NV48s_v3 Standard_NC24rs_v3 Standard_NC16as_T4_v3 Standard_NC40ads_H100_v5 Standard_ND96isr_H200_v5 Standard_NG32adms_V620_v1 Standard_NV6 Standard_NC12s_v3 Standard_NV36ads_A10_v5 Standard_NV72ads_A10_v5 Standard_NV12s_v3 Standard_NC8as_T4_v3 Standard_NC96ads_A100_v4 Standard_ND96amsr_A100_v4 Standard_NC4as_T4_v3 Standard_NV12 Standard_NV24s_v3 Standard_NV24 Standard_ND96isr_H100_v5 Standard_NG32ads_V620_v1 Standard_NV32as_v4]: spec.instanceType

Environment

  • Kubernetes version (use kubectl version):
  • OS (e.g: cat /etc/os-release):
  • Install tools:
  • Others:

Additional context

Metadata

Metadata

Labels

bugSomething isn't working

Type

No type

Projects

Status

No status

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions