Inventory

List GPU nodes available for allocation requests (GPRs) in the cluster. These nodes are managed and allocated by EGS.

List Inventory

Use this API to list Inventory.

Method and URL

GET /api/v1/inventory/list

Example Request

GET /api/v1/inventory/list
curl --location 'http://x.x.x.x:8080/api/v1/inventory/list'

Example Response

Response: Success

{
    "statusCode": 200,
    "status": "OK",
    "message": "Success",
    "data": {
        "managedNodes": [
            {
                "gpuNodeName": "gke-dev-2-gke-cluste-dev-2-gke-cluste-33478855-klfn",
                "gpuShape": "n1-highcpu-2",
                "gpuModelName": "Tesla-T4",
                "instanceType": "n1-highcpu-2",
                "clusterName": "worker-1",
                "memory": 15,
                "gpuCount": 1,
                "gpuTempThreshold": "",
                "gpuPowerThreshold": "",
                "cloudProvider": "gcp",
                "region": "",
                "nodeHealth": "Healthy",
                "gpuNodeState": "NODE_RESERVED",
                "cloud": "gcp",
                "allocation": [
                    {
                        "gprName": "gpr-b45a1d34-bf9c",
                        "sliceName": "avesha",
                        "totalGPUsAllocated": 1,
                        "allocationTimeStamp": {
                            "seconds": "1735025703",
                            "nanos": 0
                        }
                    }
                ]
            },
            {
                "gpuNodeName": "gke-dev-2-gke-cluste-dev-2-gke-cluste-33478855-kcxz",
                "gpuShape": "n1-highcpu-2",
                "gpuModelName": "Tesla-T4",
                "instanceType": "n1-highcpu-2",
                "clusterName": "worker-1",
                "memory": 15,
                "gpuCount": 1,
                "gpuTempThreshold": "",
                "gpuPowerThreshold": "",
                "cloudProvider": "gcp",
                "region": "",
                "nodeHealth": "Healthy",
                "gpuNodeState": "NODE_AVAILABLE",
                "cloud": "gcp"
            }
        ],
        "unmanagedNodes": []
    }
}

List Inventory by Slice

Use this API to list GPU nodes currently allocated to the slice.

note

Only one GPR is provisioned on a slice at any given time. Other GPRs requested will be queued and allocated one at a time based on the priority.

Method and URL

GET /api/v1/inventory?sliceName=workspace1

Example Request

GET /api/v1/inventory?sliceName=workspace1
curl --location 'http://localhost:8080/api/v1/inventory?sliceName=evening'

Example Responses

Response: Success

{
    "statusCode": 200,
    "status": "OK",
    "message": "Success",
    "data": {
        "items": [
            {
                "instanceType": "n1-highcpu-2",
                "gpuShape": "Tesla-T4",
                "gpuSlicingProfile": [],
                "memoryPerGpu": 15,
                "gpuPerNode": 1,
                "totalGpuNodes": 2,
                "clusterName": "worker-1"
            }
        ]
    }
}

Response: Not Found

{
    "statusCode": 404,
    "status": "Error",
    "message": "Not Found",
    "data": null,
    "error": {
        "errorKey": "",
        "message": "Not Found",
        "data": "{\"error\":{\"Message\":\"Error occurred while fetching the inventory details\",\"Status\":404,\"DetailedError\":{\"Errormessage\":\"sliceconfigs.controller.kubeslice.io \\\"workspace2\\\" not found\",\"Reason\":\"NotFound\",\"Details\":\"\"},\"StatusCode\":404,\"ReferenceDocumentation\":\"NA\"}}"
    }
}