azure-sdk-for-go: ML Accounts List function giving error

I am trying to call machine learning experimentation accounts List function. I am using the “github.com/Azure/azure-sdk-for-go/services/machinelearning/mgmt/2017-05-01-preview/experimentation” version.

When I make a call like this:

....
result, err := client.List(context.Background())
....

I get the following error:

experimentation.AccountsClient#List: Failure responding to request: StatusCode=200 -- Original Error: Error occurred unmarshalling JSON - Error = 'json: cannot unmarshal string into Go struct field AccountListResult.value of type experimentation.StorageAccountProperties'

I can see in the detailed JSON it returns that storage account is a string ID, however, in models section, Storage Account is a struct with two fields: ID and access key.

About this issue

Original URL
State: closed
Created 6 years ago
Comments: 18 (10 by maintainers)

Most upvoted comments

I see. Here’s the swagger for Studio workspace: https://github.com/Azure/azure-rest-api-specs/blob/master/specification/machinelearning/resource-manager/Microsoft.MachineLearning/stable/2016-04-01/workspaces.json

I will need to check about model management.

FYI: @hning86 @gokhanuluderya-msft

ragargms on May 8, 2018

Hello,

You need a team account in order to create a workspace. This is because workspaces are nested resources under accounts. As for the requirements for key vault, vso and storage account, we had those requirements earlier but key vault and vso got dropped soon after. We will update the model soon. However, you will still need to provide storage account and key for your storage account when you create an experimentation account. You can get the key for storage account from the Azure portal. You can also install machine learning workbench, https://docs.microsoft.com/en-us/azure/machine-learning/service/quickstart-installation which gives you a CLI interface to create accounts/workspaces etc.

For now, you can ignore the error message, we will fix the models soon but in the meantime I highly encourage you to try workbench

ragargms on May 8, 2018