New compute wizard fails with MinNodeCount must be specified.

Model Builder Version: 16.14.0.2255902
Visual Studion Version: 17.4.1

Bug description
Receive error when trying to create compute resources for an object detection project in Visual Studio.

ScaleSettings.MaxNodeCount and ScaleSettings.MinNodeCount must be specified for create compute.
Status: 400 (ScaleSettings.MaxNodeCount and ScaleSettings.MinNodeCount must be specified for create compute.)
ErrorCode: UserError

Steps to Reproduce

Create a new .mbconfig file
Choose Object Detection scenario
On the Environment tab, click the Set up workspace button
Selected Visual Studio Enterprise Subscription
Create a new Machine Learning Workspace name
Click the New... button
Set the Compute name
St the Compute Size to Standard_NC4as_T4_v3
Set Compute priority to Low priority
Click the Create button

Expected Experience
I would expect that the compute resources would be created, and I would be returned to the previous dialog in the wizard.

Actual Experience
I receive the following error

ScaleSettings.MaxNodeCount and ScaleSettings.MinNodeCount must be specified for create compute.
Status: 400 (ScaleSettings.MaxNodeCount and ScaleSettings.MinNodeCount must be specified for create compute.)
ErrorCode: UserError

Content:
{
  "error": {
    "code": "UserError",
    "severity": null,
    "message": "ScaleSettings.MaxNodeCount and ScaleSettings.MinNodeCount must be specified for create compute.",
    "messageFormat": null,
    "messageParameters": null,
    "referenceCode": null,
    "detailsUri": null,
    "target": null,
    "details": [],
    "innerError": null,
    "debugInfo": null,
    "additionalInfo": null
  },
  "correlation": {
    "operation": "af0c4e14b071bf1ad124097cc8138c77",
    "request": "c6a058345902b3bf"
  },
  "environment": "northcentralus",
  "location": "northcentralus",
  "time": "2022-11-18T16:11:32.7653884+00:00",
  "componentName": "machinelearningcompute"
}

Headers:
Pragma: no-cache
x-ms-ratelimit-remaining-subscription-writes: REDACTED
Request-Context: REDACTED
x-ms-response-type: REDACTED
Strict-Transport-Security: REDACTED
X-Content-Type-Options: REDACTED
x-aml-cluster: REDACTED
x-request-time: REDACTED
x-ms-request-id: 670f5fe2-52f6-47a5-9e71-1177a8af2560
x-ms-correlation-request-id: REDACTED
x-ms-routing-request-id: REDACTED
Content-Length: 673
Cache-Control: no-cache
Content-Type: application/json; charset=utf-8
Date: Fri, 18 Nov 2022 16:11:32 GMT
Expires: -1

at Azure.ResourceManager.MachineLearning.ComputeRestOperations.<CreateOrUpdateAsync>d__12.MoveNext()
--- End of stack trace from previous location where exception was thrown ---
   at System.Runtime.ExceptionServices.ExceptionDispatchInfo.Throw()
   at System.Runtime.CompilerServices.TaskAwaiter.HandleNonSuccessAndDebuggerNotification(Task task)
   at Azure.ResourceManager.MachineLearning.MachineLearningComputeCollection.<CreateOrUpdateAsync>d__5.MoveNext()
--- End of stack trace from previous location where exception was thrown ---
   at System.Runtime.ExceptionServices.ExceptionDispatchInfo.Throw()
   at System.Runtime.CompilerServices.TaskAwaiter.HandleNonSuccessAndDebuggerNotification(Task task)
   at Microsoft.ML.ModelBuilder.RemoteResourceManager.AzureResourceManager.<CreateAMLComputeAsync>d__11.MoveNext()
--- End of stack trace from previous location where exception was thrown ---
   at System.Runtime.ExceptionServices.ExceptionDispatchInfo.Throw()
   at System.Runtime.CompilerServices.TaskAwaiter.HandleNonSuccessAndDebuggerNotification(Task task)
   at Microsoft.ML.ModelBuilder.AzureAttach.NewComputeDataContext.<CreateComputeAsync>d__23.MoveNext()

Additional Context

I am running Visual Studio in a Parallels VM on a MacBook Pro.

@jaydeboer

@jaydeboer, sorry for hitting the issue, we will investigate it.

Having the same problem with Standard_NC6 image. Visual Studio 2022, up to date.

`ScaleSettings.MaxNodeCount and ScaleSettings.MinNodeCount must be specified for create compute.
Status: 400 (ScaleSettings.MaxNodeCount and ScaleSettings.MinNodeCount must be specified for create compute.)
ErrorCode: UserError

Content:
{
"error": {
"code": "UserError",
"severity": null,
"message": "ScaleSettings.MaxNodeCount and ScaleSettings.MinNodeCount must be specified for create compute.",
"messageFormat": null,
"messageParameters": null,
"referenceCode": null,
"detailsUri": null,
"target": null,
"details": [],
"innerError": null,
"debugInfo": null,
"additionalInfo": null
},
"correlation": {
"operation": "0c60e2ac1be04499761abfef25beb033",
"request": "997aeb1b262dfb50"
},
"environment": "westeurope",
"location": "westeurope",
"time": "2023-01-06T10:54:55.7634809+00:00",
"componentName": "machinelearningcompute"
}

Headers:
Pragma: no-cache
x-ms-ratelimit-remaining-subscription-writes: REDACTED
Request-Context: REDACTED
x-ms-response-type: REDACTED
Strict-Transport-Security: REDACTED
X-Content-Type-Options: REDACTED
x-aml-cluster: REDACTED
x-request-time: REDACTED
x-ms-request-id: 1b4fef80-045d-40fe-bbc6-7013444f3bdf
x-ms-correlation-request-id: REDACTED
x-ms-routing-request-id: REDACTED
Content-Length: 665
Cache-Control: no-cache
Content-Type: application/json; charset=utf-8
Date: Fri, 06 Jan 2023 10:54:55 GMT
Expires: -1
`

Same issue here:

@luisquintanilla

@luisquintanilla We're getting multiple reports of this problem. We should investigate, I've added this to the March milestone.

@beccamc

@beccamc sounds good. Thanks.Bumped it to Pri0

Work around for those struggling with this. You may be able to create a compute in the Azure ML Studio - https://ml.azure.com/

Go to Compute -> Compute Cluster -> New.

Any of these compute types will work....

NC T4 v3 VM series:

"STANDARD_NC4AS_T4_V3",
"STANDARD_NC16As_T4_V3"
"STANDARD_NC64As_T4_V3"
"STANDARD_NC24RS_V3"

NCv3-series:

"STANDARD_NC6S_V3"
"STANDARD_NC12S_V3"
"STANDARD_NC24S_V3"
"STANDARD_NC24RS_V3*"

Once you create the compute it should be available via Model Builder.

heyyy, this worked for me! Created a "Standard_NC4as_T4_v3". Just need to select it in the Model Builder UI now!

@JakeRadMSFT

@JakeRadMSFT You probably want to fix this.

Wanted to +1 this as I am encountering the same problem while following the tutorial from here: https://learn.microsoft.com/en-us/azure/machine-learning/tutorial-train-deploy-image-classification-model-vscode

The bug has been fixed on the latest main build 17.14.4.2318101.

Work around for those struggling with this. You may be able to create a compute in the Azure ML Studio - https://ml.azure.com/

Go to Compute -> Compute Cluster -> New.

Any of these compute types will work....

NC T4 v3 VM series:

"STANDARD_NC4AS_T4_V3",

"STANDARD_NC16As_T4_V3"

"STANDARD_NC64As_T4_V3"

"STANDARD_NC24RS_V3"

NCv3-series:

"STANDARD_NC6S_V3"

"STANDARD_NC12S_V3"

"STANDARD_NC24S_V3"

"STANDARD_NC24RS_V3*"

Once you create the compute it should be available via Model Builder.

I have created a STANDARD_NC64As_T4_V3, but it does not appear when selecting Compute. Only the sizes that are displayed in the drop-down menu when creating Compute are shown.
ML Model Builder Version : 17.14.4.2312404

Navigation Menu

Repository navigation

New compute wizard fails with MinNodeCount must be specified. #2386

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

New compute wizard fails with MinNodeCount must be specified. #2386

Description

Activity

zewditu commented on Nov 28, 2022

Jenscaasen commented on Jan 6, 2023

ktravelet commented on Feb 16, 2023

beccamc commented on Feb 16, 2023

luisquintanilla commented on Feb 16, 2023

beccamc commented on Feb 16, 2023

ktravelet commented on Feb 16, 2023

beccamc commented on Mar 3, 2023

slefcourt27 commented on Mar 12, 2023

v-Hailishi commented on Apr 3, 2023

johnseed commented on Apr 6, 2023

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions