AttributeError: 'BlobServiceClient' object has no attribute '_to_rest_object'

I was trying to add a blob storage connection to MLClient and it gives me an attribute error.

What I have:

client=MLClient(
    DefaultAzureCredential(), 
    subscription_id=sub_id,
    resource_group_name=group_id,
    workspace_name=workspace
    )

blob_service_client = BlobServiceClient(account_url=account_url,
                                        credential=credential)

client.connections.create_or_update(blob_service_client)

Output:

rest_workspace_connection = workspace_connection._to_rest_object()
AttributeError: 'BlobServiceClient' object has no attribute '_to_rest_object'

NOTE: I checked everything. All the links to the storage and credentials are correct, so it should be working fine.

Is there any way I can make a connection with blob storage without the lobServiceClient class? I believe something is wrong with it.

Is there any way I can make a connection with blob storage without the blobServiceClient class? I believe something is wrong with it.

You can use the below code that will create connection in azure machine learning using Azure python SDK.

Code:

from azure.ai.ml import MLClient
from azure.ai.ml.entities import AzureBlobStoreConnection,SasTokenConfiguration
from azure.identity import DefaultAzureCredential

# Replace with your Azure details
subscription_id = "xxx"
resource_group = "xxx"
workspace_name = "xxx"

# Storage details
connection_name = "my_blob_store"
account_name = "venkat906"
container_name = "test"
url="xxxx"

# Initialize MLClient with DefaultAzureCredential removed
ml_client = MLClient(
    credential=DefaultAzureCredential(),  # Optionally, you can try removing this to avoid conflicts
    subscription_id=subscription_id,
    resource_group_name=resource_group,
    workspace_name=workspace_name,
)

# Create AzureBlobStoreConnection entity with AccountKeyConfiguration
blob_connection = AzureBlobStoreConnection(
    name=connection_name,
    url=url,
    container_name=container_name,
    account_name=account_name,
    credentials=SasTokenConfiguration(sas_token="sv=2022-11-02&ss=bfqt&srt=sco&sp=rwdlacupiytfx&se=2025-01-16T14:36:46Z&st=2025-01-16T06:36:46Z&spr=https&sig=redacted")
)

# Create or update the connection
try:
    ml_client.connections.create_or_update(blob_connection)
    print(f"Blob storage connection '{connection_name}' created successfully.")
except Exception as e:
    print(f"Failed to create connection: {e}")

connection = ml_client.connections.get(name=connection_name)
print(f"Connection '{connection.name}' created successfully.")
print(f"Type: {connection.type}")
print(f"Target URL: {connection.target}") 

Output:

Blob storage connection 'my_blob_store' created successfully.
Connection 'my_blob_store' created successfully.
Type: azure_blob
Target URL: /subscriptions/cxx/resourceGroups/xx/providers/Microsoft.Storage/storageAccounts/venkat906/blobServices/default

Portal:

Reference: azure.ai.ml.entities.AzureBlobStoreConnection class | Microsoft Learn