My python code uses AzureCliCredential() function, but it is giving error in running in Synapse/ADF/Databricks notebook .

Utsav Mori 20

My python code uses AzureCliCredential() function, but it is giving error in running in Synapse Workspace that your Azure Cli command not found on path .

I have tried using other function like DefaultAzureCredentials(), ClientSecretCredentials () also

ClientSecretCredential(
     tenant_id="***",
     client_id="***",
     client_secret="***"
 )

But above two functions are giving error in my local env too .

The only function that is working without any error is AzureCliCredentials() and I want my python code containing AzureCliCredentials function to get deploy on Synapse/ADF/Databricks and run successfully without showing error AzureCli Path not found, Please assist me for the same .

phemanth 8,645 Reputation points Microsoft Vendor

2024-07-19T09:15:27.3033333+00:00
@Utsav Mori
Thanks for the question and using MS Q&A platform

It seems like you're facing an issue with using AzureCliCredential() in Synapse Workspace, which is complaining about the Azure Cli command not being found on the path.

You've tried using DefaultAzureCredentials() and ClientSecretCredentials() as well, but they're giving errors in your local environment too.

The reason AzureCliCredential() is working without errors is that it uses the Azure CLI to authenticate, which is not the case with the other two credentials.

To deploy your Python code containing AzureCliCredentials function to Synapse/ADF/Databricks and run it successfully without the AzureCli Path not found error, you need to ensure that the Azure CLI is installed and configured properly in your Synapse Workspace environment.

Here's an example of how you can use AzureCliCredential():

from azure.identity import AzureCliCredential

I have tried my end and it has worked for me in synapse note book

with

ClientSecretCredentials() is also working for me as shown below

and with

AzureCliCredential()

However, since you want to deploy your code to Synapse/ADF/Databricks, you might need to consider using DefaultAzureCredential() instead, as they are more suitable for managed environments.

To use the DefaultAzureCredential function, follow these steps:

Step 1: Install the azure-identity library

If you haven't already, install the azure-identity library using pip:

pip install azure-identity

Step 2: Import the DefaultAzureCredential function

In your Python code, import the DefaultAzureCredential function:

from azure.identity import DefaultAzureCredential

Step 3: Create a credential object

Create a credential object using the DefaultAzureCredential function:

credential = DefaultAzureCredential()

This will create a credential object that will try to authenticate using the methods mentioned above.

Step 4: Use the credential object

You can now use the credential object to authenticate with Azure services, such as Azure Storage, Azure Key Vault, or Azure Data Lake Storage.

For example, to use the credential object with Azure Storage:

from azure.storage.blob import BlobServiceClient blob_service_client = BlobServiceClient("https://<account_name>.blob.core.windows.net/", credential=credential)

Replace <account_name> link with the name of your Azure Storage account.

If you're still facing issues, please provide more details about your environment and error logs, and I'll be happy to help you further.

Hope this helps. Do let us know if you any further queries.
Deleted

This comment has been deleted due to a violation of our Code of Conduct. The comment was manually reported or identified through automated detection before action was taken. Please refer to our Code of Conduct for more information.

Utsav Mori 20

Hi ,

I am trying the same and it is not giving error till this point i.e:

from azure.identity import AzureCliCredential
# from azure.identity import ClientSecretCredential
from azure.identity import DefaultAzureCredential
from pyapacheatlas.core import PurviewClient
import json
keywords = '*'
# cred = AzureCliCredential()
# print(cred)
cred = DefaultAzureCredential()
print(cred)
client1 = PurviewCatalogClient(endpoint="account-name", credential=cred, logging_enable=True)
print(client1)

but after writing the below line it is giving me DefaultAzureCredentials error, Below code is giving me error that is this response line:

body_input = {
    "keywords": keywords,
    "limit": 1,
    # "offset":l,
    #"filter": {"assetType": ""}
}

response = client1.discovery.query(search_request=body_input)

this above response line is giving me error:

User's image

Note : I can't use ClientSecretCredentials() because I can't access client_secret due to client side restriction.

Below is my overall code for your reference:

from azure.purview.catalog import PurviewCatalogClient
from azure.identity import AzureCliCredential
# from azure.identity import ClientSecretCredential
from azure.identity import DefaultAzureCredential
from pyapacheatlas.core import PurviewClient
import json

keywords = '*'

# cred = AzureCliCredential()
# print(cred)
cred = DefaultAzureCredential()
print(cred)
client1 = PurviewCatalogClient(endpoint="account-name", credential=cred, logging_enable=True)
print(client1)

# client = PurviewClient(
#      account_name="",
#      authentication = cred
# )
print(client)
body_input = {
    "keywords": keywords,
    "limit": 1,
    # "offset":l,
    #"filter": {"assetType": ""}
}

response = client1.discovery.query(search_request=body_input)
print(response)

Share via

My python code uses AzureCliCredential() function, but it is giving error in running in Synapse/ADF/Databricks notebook .