MANAGED_IDENTITY_CLIENT_ID to client_id

Author: Pavan-Microsoft

infra/scripts/index_scripts/01_create_search_index.py

@@ -18,6 +18,7 @@
 
 # === Configuration ===
 KEY_VAULT_NAME = 'kv_to-be-replaced'
+MANAGED_IDENTITY_CLIENT_ID = 'mici_to-be-replaced'
 INDEX_NAME = "call_transcripts_index"
 
 
@@ -32,7 +33,7 @@ def get_secrets_from_kv(secret_name: str) -> str:
     Returns:
         str: The secret value.
     """
-    kv_credential = ManagedIdentityCredential()
+    kv_credential = ManagedIdentityCredential(client_id=MANAGED_IDENTITY_CLIENT_ID)
     secret_client = SecretClient(
         vault_url=f"https://{KEY_VAULT_NAME}.vault.azure.net/",
         credential=kv_credential
@@ -48,7 +49,7 @@ def create_search_index():
     - Semantic search using prioritized fields
     """
     # Shared credential
-    credential = ManagedIdentityCredential()
+    credential = ManagedIdentityCredential(client_id=MANAGED_IDENTITY_CLIENT_ID)
 
     # Retrieve secrets from Key Vault
     search_endpoint = get_secrets_from_kv("AZURE-SEARCH-ENDPOINT")
@@ -120,4 +121,4 @@ def create_search_index():
     print(f"Search index '{result.name}' created or updated successfully.")
 
 
-create_search_index()
+create_search_index()

infra/scripts/index_scripts/02_create_cu_template_audio.py

@@ -9,6 +9,7 @@
 
 # === Configuration ===
 KEY_VAULT_NAME = 'kv_to-be-replaced'
+MANAGED_IDENTITY_CLIENT_ID = 'mici_to-be-replaced'
 AZURE_AI_API_VERSION = "2024-12-01-preview"
 ANALYZER_ID = "ckm-audio"
 ANALYZER_TEMPLATE_FILE = 'ckm-analyzer_config_audio.json'
@@ -26,7 +27,7 @@ def get_secrets_from_kv(secret_name: str, vault_name: str) -> str:
     Returns:
         str: The value of the secret.
     """
-    kv_credential = ManagedIdentityCredential()
+    kv_credential = ManagedIdentityCredential(client_id=MANAGED_IDENTITY_CLIENT_ID)
     secret_client = SecretClient(
         vault_url=f"https://{vault_name}.vault.azure.net/",
         credential=kv_credential
@@ -38,7 +39,7 @@ def get_secrets_from_kv(secret_name: str, vault_name: str) -> str:
 # Fetch endpoint from Key Vault
 endpoint = get_secrets_from_kv("AZURE-OPENAI-CU-ENDPOINT", KEY_VAULT_NAME)
 
-credential = ManagedIdentityCredential()
+credential = ManagedIdentityCredential(client_id=MANAGED_IDENTITY_CLIENT_ID)
 # Initialize Content Understanding Client
 token_provider = get_bearer_token_provider(credential, "https://cognitiveservices.azure.com/.default")
 client = AzureContentUnderstandingClient(
@@ -49,4 +50,4 @@ def get_secrets_from_kv(secret_name: str, vault_name: str) -> str:
 
 # Create Analyzer
 response = client.begin_create_analyzer(ANALYZER_ID, analyzer_template_path=ANALYZER_TEMPLATE_FILE)
-result = client.poll_result(response)
+result = client.poll_result(response)

infra/scripts/index_scripts/02_create_cu_template_text.py

@@ -9,6 +9,7 @@
 
 # === Configuration ===
 KEY_VAULT_NAME = 'kv_to-be-replaced'
+MANAGED_IDENTITY_CLIENT_ID = 'mici_to-be-replaced'
 AZURE_AI_API_VERSION = "2024-12-01-preview"
 ANALYZER_ID = "ckm-json"
 ANALYZER_TEMPLATE_FILE = 'ckm-analyzer_config_text.json'
@@ -19,7 +20,7 @@ def get_secret(secret_name: str, vault_name: str) -> str:
     """
     Retrieve a secret value from Azure Key Vault.
     """
-    kv_credential = ManagedIdentityCredential()
+    kv_credential = ManagedIdentityCredential(client_id=MANAGED_IDENTITY_CLIENT_ID)
     secret_client = SecretClient(vault_url=f"https://{vault_name}.vault.azure.net/", credential=kv_credential)
     return secret_client.get_secret(secret_name).value
 
@@ -29,7 +30,7 @@ def get_secret(secret_name: str, vault_name: str) -> str:
 # Get endpoint from Key Vault
 endpoint = get_secret("AZURE-OPENAI-CU-ENDPOINT", KEY_VAULT_NAME)
 
-credential = ManagedIdentityCredential()
+credential = ManagedIdentityCredential(client_id=MANAGED_IDENTITY_CLIENT_ID)
 # Initialize Content Understanding Client
 token_provider = get_bearer_token_provider(credential, "https://cognitiveservices.azure.com/.default")
 client = AzureContentUnderstandingClient(
@@ -40,4 +41,4 @@ def get_secret(secret_name: str, vault_name: str) -> str:
 
 # Create Analyzer
 response = client.begin_create_analyzer(ANALYZER_ID, analyzer_template_path=ANALYZER_TEMPLATE_FILE)
-result = client.poll_result(response)
+result = client.poll_result(response)

infra/scripts/index_scripts/03_cu_process_data_text.py

@@ -15,13 +15,14 @@
 
 # Constants and configuration
 KEY_VAULT_NAME = 'kv_to-be-replaced'
+MANAGED_IDENTITY_CLIENT_ID = 'mici_to-be-replaced'
 FILE_SYSTEM_CLIENT_NAME = "data"
 DIRECTORY = 'call_transcripts'
 AUDIO_DIRECTORY = 'audiodata'
 INDEX_NAME = "call_transcripts_index"
 
 def get_secrets_from_kv(kv_name, secret_name):
-    kv_credential = ()
+    kv_credential = ManagedIdentityCredential(client_id=MANAGED_IDENTITY_CLIENT_ID)
     secret_client = SecretClient(vault_url=f"https://{kv_name}.vault.azure.net/", credential=kv_credential)
     return secret_client.get_secret(secret_name).value
 
@@ -39,15 +40,15 @@ def get_secrets_from_kv(kv_name, secret_name):
 
 # Azure DataLake setup
 account_url = f"https://{account_name}.dfs.core.windows.net"
-credential = ManagedIdentityCredential()
+credential = ManagedIdentityCredential(client_id=MANAGED_IDENTITY_CLIENT_ID)
 service_client = DataLakeServiceClient(account_url, credential=credential, api_version='2023-01-03')
 file_system_client = service_client.get_file_system_client(FILE_SYSTEM_CLIENT_NAME)
 directory_name = DIRECTORY
 paths = list(file_system_client.get_paths(path=directory_name))
 print("Azure DataLake setup complete.")
 
 # Azure Search setup
-search_credential = ManagedIdentityCredential()
+search_credential = ManagedIdentityCredential(client_id=MANAGED_IDENTITY_CLIENT_ID)
 search_client = SearchClient(search_endpoint, INDEX_NAME, search_credential)
 index_client = SearchIndexClient(endpoint=search_endpoint, credential=search_credential)
 print("Azure Search setup complete.")
@@ -64,7 +65,7 @@ def get_secrets_from_kv(kv_name, secret_name):
 
 
 # Content Understanding client
-cu_credential = ManagedIdentityCredential()
+cu_credential = ManagedIdentityCredential(client_id=MANAGED_IDENTITY_CLIENT_ID)
 cu_token_provider = get_bearer_token_provider(cu_credential, "https://cognitiveservices.azure.com/.default")
 cu_client = AzureContentUnderstandingClient(
     endpoint=azure_ai_endpoint,
@@ -78,7 +79,7 @@ def get_secrets_from_kv(kv_name, secret_name):
 def get_embeddings(text: str, openai_api_base, openai_api_version):
     model_id = "text-embedding-ada-002"
     token_provider = get_bearer_token_provider(
-        ManagedIdentityCredential(),
+        ManagedIdentityCredential(client_id=MANAGED_IDENTITY_CLIENT_ID),
         "https://cognitiveservices.azure.com/.default"
     )
     client = AzureOpenAI(
@@ -289,7 +290,7 @@ def call_gpt4(topics_str1, client):
     return json.loads(res.replace("```json", '').replace("```", ''))
 
 token_provider = get_bearer_token_provider(
-    ManagedIdentityCredential(),
+    ManagedIdentityCredential(client_id=MANAGED_IDENTITY_CLIENT_ID),
     "https://cognitiveservices.azure.com/.default"
 )
 openai_client = AzureOpenAI(

infra/scripts/index_scripts/04_cu_process_data_new_data.py

@@ -29,13 +29,14 @@
 
 # Constants and configuration
 KEY_VAULT_NAME = 'kv_to-be-replaced'
+MANAGED_IDENTITY_CLIENT_ID = 'mici_to-be-replaced'
 FILE_SYSTEM_CLIENT_NAME = "data"
 DIRECTORY = 'custom_transcripts'
 AUDIO_DIRECTORY = 'custom_audiodata'
 INDEX_NAME = "call_transcripts_index"
 
 def get_secrets_from_kv(kv_name, secret_name):
-    kv_credential = ManagedIdentityCredential()
+    kv_credential = ManagedIdentityCredential(client_id=MANAGED_IDENTITY_CLIENT_ID)
     secret_client = SecretClient(vault_url=f"https://{kv_name}.vault.azure.net/", credential=kv_credential)
     return secret_client.get_secret(secret_name).value
 
@@ -54,15 +55,15 @@ def get_secrets_from_kv(kv_name, secret_name):
 
 # Azure DataLake setup
 account_url = f"https://{account_name}.dfs.core.windows.net"
-credential = ManagedIdentityCredential()
+credential = ManagedIdentityCredential(client_id=MANAGED_IDENTITY_CLIENT_ID)
 service_client = DataLakeServiceClient(account_url, credential=credential, api_version='2023-01-03')
 file_system_client = service_client.get_file_system_client(FILE_SYSTEM_CLIENT_NAME)
 directory_name = DIRECTORY
 paths = list(file_system_client.get_paths(path=directory_name))
 print("Azure DataLake setup complete.")
 
 # Azure Search setup
-search_credential = ManagedIdentityCredential()
+search_credential = ManagedIdentityCredential(client_id=MANAGED_IDENTITY_CLIENT_ID)
 search_client = SearchClient(search_endpoint, INDEX_NAME, search_credential)
 index_client = SearchIndexClient(endpoint=search_endpoint, credential=search_credential)
 print("Azure Search setup complete.")
@@ -155,7 +156,7 @@ def create_search_index():
 print("SQL Server connection established.")
 
 # Content Understanding client
-cu_credential = ManagedIdentityCredential()
+cu_credential = ManagedIdentityCredential(client_id=MANAGED_IDENTITY_CLIENT_ID)
 cu_token_provider = get_bearer_token_provider(cu_credential, "https://cognitiveservices.azure.com/.default")
 cu_client = AzureContentUnderstandingClient(
     endpoint=azure_ai_endpoint,
@@ -168,7 +169,7 @@ def create_search_index():
 def get_embeddings(text: str, openai_api_base, openai_api_version):
     model_id = "text-embedding-ada-002"
     token_provider = get_bearer_token_provider(
-        ManagedIdentityCredential(),
+        ManagedIdentityCredential(client_id=MANAGED_IDENTITY_CLIENT_ID),
         "https://cognitiveservices.azure.com/.default"
     )
     client = AzureOpenAI(
@@ -403,7 +404,7 @@ def call_gpt4(topics_str1, client):
     return json.loads(res.replace("```json", '').replace("```", ''))
 
 token_provider = get_bearer_token_provider(
-    ManagedIdentityCredential(),
+    ManagedIdentityCredential(client_id=MANAGED_IDENTITY_CLIENT_ID),
     "https://cognitiveservices.azure.com/.default"
 )
 openai_client = AzureOpenAI(