Revert 'Adds X-Request-LightspeedUser to WCA requests (#1667)' (#1669)

Author: hasys

ansible_ai_connect/ai/api/model_pipelines/tests/test_wca_client.py

@@ -58,7 +58,6 @@
 from ansible_ai_connect.ai.api.model_pipelines.tests import mock_pipeline_config
 from ansible_ai_connect.ai.api.model_pipelines.wca.pipelines_base import (
     WCA_REQUEST_ID_HEADER,
-    WCA_REQUEST_USER_UUID_HEADER,
     ibm_cloud_identity_token_hist,
     ibm_cloud_identity_token_retry_counter,
     wca_codegen_hist,
@@ -847,7 +846,6 @@ def _do_inference(
     ):
         model_id = "zavala"
         api_key = "abc123"
-        user_uuid = str(uuid.uuid4())
         context = ""
         prompt = prompt if prompt else "- name: install ffmpeg on Red Hat Enterprise Linux"
 
@@ -876,14 +874,13 @@ def _do_inference(
         response = MockResponse(
             json=predictions,
             status_code=200,
-            headers={WCA_REQUEST_ID_HEADER: request_id, WCA_REQUEST_USER_UUID_HEADER: user_uuid},
+            headers={WCA_REQUEST_ID_HEADER: request_id},
         )
 
         requestHeaders = {
             "Content-Type": "application/json",
             "Authorization": f"Bearer {token['access_token']}",
             WCA_REQUEST_ID_HEADER: suggestion_id,
-            WCA_REQUEST_USER_UUID_HEADER: user_uuid,
         }
 
         model_client = WCASaaSCompletionsPipeline(self.config)
@@ -892,13 +889,9 @@ def _do_inference(
         model_client.get_model_id = Mock(return_value=model_id)
         model_client.get_api_key = Mock(return_value=api_key)
 
-        mock_request = Mock()
-        mock_request.user = Mock()
-        mock_request.user.uuid = user_uuid
-
         result = model_client.invoke(
             CompletionsParameters.init(
-                request=mock_request,
+                request=Mock(),
                 model_input=model_input,
                 model_id=model_id,
                 suggestion_id=suggestion_id,
@@ -1492,7 +1485,6 @@ def test_get_model_id_without_setting(self):
 class TestWCAOnPremCodegen(WisdomServiceLogAwareTestCase):
     prompt = "- name: install ffmpeg on Red Hat Enterprise Linux"
     suggestion_id = "suggestion_id"
-    user_uuid = str(uuid.uuid4())
     token = base64.b64encode(bytes("username:12345", "ascii")).decode("ascii")
     codegen_data = {
         "model_id": "model-name",
@@ -1501,7 +1493,6 @@ class TestWCAOnPremCodegen(WisdomServiceLogAwareTestCase):
     request_headers = {
         "Authorization": f"ZenApiKey {token}",
         WCA_REQUEST_ID_HEADER: suggestion_id,
-        WCA_REQUEST_USER_UUID_HEADER: user_uuid,
     }
     model_input = {
         "instances": [
@@ -1528,13 +1519,9 @@ def setUp(self):
         self.model_client.session.post = Mock(return_value=MockResponse(json={}, status_code=200))
 
     def test_headers(self):
-        mock_request = Mock()
-        mock_request.user = Mock()
-        mock_request.user.uuid = self.user_uuid
-
         self.model_client.invoke(
             CompletionsParameters.init(
-                request=mock_request, model_input=self.model_input, suggestion_id=self.suggestion_id
+                request=Mock(), model_input=self.model_input, suggestion_id=self.suggestion_id
             ),
         )
         self.model_client.session.post.assert_called_once_with(
@@ -1546,13 +1533,10 @@ def test_headers(self):
         )
 
     def test_disabled_model_server_ssl(self):
-        mock_request = Mock()
-        mock_request.user = Mock()
-        mock_request.user.uuid = self.user_uuid
         self.config.verify_ssl = False
         self.model_client.invoke(
             CompletionsParameters.init(
-                request=mock_request, model_input=self.model_input, suggestion_id=self.suggestion_id
+                request=Mock(), model_input=self.model_input, suggestion_id=self.suggestion_id
             ),
         )
         self.model_client.session.post.assert_called_once_with(

ansible_ai_connect/ai/api/model_pipelines/wca/pipelines_base.py

@@ -81,8 +81,6 @@
 
 WCA_REQUEST_ID_HEADER = "X-Request-ID"
 
-WCA_REQUEST_USER_UUID_HEADER = "X-Request-LightspeedUser"
-
 # from django_prometheus.middleware.DEFAULT_LATENCY_BUCKETS
 DEFAULT_LATENCY_BUCKETS = (
     0.01,
@@ -245,20 +243,6 @@ class WCABasePipeline(
     def __init__(self, config: WCA_PIPELINE_CONFIGURATION):
         super().__init__(config=config)
 
-    def _prepare_request_headers(
-        self, request_user: Optional[User], api_key: str, identifier: Optional[str]
-    ) -> dict[str, Optional[str]]:
-        """
-        Helper method to extract user UUID and get request headers.
-        """
-        lightspeed_user_uuid_str: Optional[str] = None
-        if request_user and hasattr(request_user, "uuid"):
-            lightspeed_user_uuid_str = str(request_user.uuid)
-
-        return self.get_request_headers(
-            api_key, identifier, lightspeed_user_uuid=lightspeed_user_uuid_str
-        )
-
     @staticmethod
     def log_backoff_exception(details):
         _, exc, _ = sys.exc_info()
@@ -300,7 +284,7 @@ def on_backoff_explain_role(details):
 
     @abstractmethod
     def get_request_headers(
-        self, api_key: str, identifier: Optional[str], lightspeed_user_uuid: Optional[str] = None
+        self, api_key: str, identifier: Optional[str]
     ) -> dict[str, Optional[str]]:
         raise NotImplementedError
 
@@ -334,12 +318,7 @@ def invoke(self, params: CompletionsParameters) -> CompletionsResponse:
         try:
             api_key = self.get_api_key(request.user)
             model_id = self.get_model_id(request.user, model_id)
-
-            headers = self._prepare_request_headers(request.user, api_key, suggestion_id)
-
-            result = self.infer_from_parameters(
-                api_key, model_id, context, prompt, suggestion_id, headers
-            )
+            result = self.infer_from_parameters(api_key, model_id, context, prompt, suggestion_id)
 
             response = result.json()
             response["model_id"] = model_id
@@ -349,15 +328,14 @@ def invoke(self, params: CompletionsParameters) -> CompletionsResponse:
         except requests.exceptions.Timeout:
             raise ModelTimeoutError(model_id=model_id)
 
-    def infer_from_parameters(
-        self, api_key, model_id, context, prompt, suggestion_id=None, headers=None
-    ):
+    def infer_from_parameters(self, api_key, model_id, context, prompt, suggestion_id=None):
         data = {
             "model_id": model_id,
             "prompt": f"{context}{prompt}",
         }
         logger.debug(f"Inference API request payload: {json.dumps(data)}")
 
+        headers = self.get_request_headers(api_key, suggestion_id)
         task_count = len(get_task_names_from_prompt(prompt))
         prediction_url = f"{self.config.inference_url}/v1/wca/codegen/ansible"
 
@@ -493,8 +471,7 @@ def invoke(self, params: PlaybookGenerationParameters) -> PlaybookGenerationResp
         api_key = self.get_api_key(request.user)
         model_id = self.get_model_id(request.user, model_id)
 
-        headers = self._prepare_request_headers(request.user, api_key, generation_id)
-
+        headers = self.get_request_headers(api_key, generation_id)
         data = {
             "model_id": model_id,
             "text": text,
@@ -576,8 +553,7 @@ def invoke(self, params: RoleGenerationParameters) -> RoleGenerationResponse:
         api_key = self.get_api_key(request.user)
         model_id = self.get_model_id(request.user, model_id)
 
-        headers = self._prepare_request_headers(request.user, api_key, generation_id)
-
+        headers = self.get_request_headers(api_key, generation_id)
         data = {
             "model_id": model_id,
             "text": text,
@@ -658,8 +634,7 @@ def invoke(self, params: PlaybookExplanationParameters) -> PlaybookExplanationRe
         api_key = self.get_api_key(request.user)
         model_id = self.get_model_id(request.user, model_id)
 
-        headers = self._prepare_request_headers(request.user, api_key, explanation_id)
-
+        headers = self.get_request_headers(api_key, explanation_id)
         data = {
             "model_id": model_id,
             "playbook": content,
@@ -721,8 +696,7 @@ def invoke(self, params: RoleExplanationParameters) -> RoleExplanationResponse:
         api_key = self.get_api_key(request.user)
         model_id = self.get_model_id(request.user, model_id)
 
-        headers = self._prepare_request_headers(request.user, api_key, explanation_id)
-
+        headers = self.get_request_headers(api_key, explanation_id)
         data = {
             "role_name": params.role_name,
             "model_id": model_id,

ansible_ai_connect/ai/api/model_pipelines/wca/pipelines_onprem.py

@@ -47,7 +47,6 @@
 )
 from ansible_ai_connect.ai.api.model_pipelines.wca.pipelines_base import (
     WCA_REQUEST_ID_HEADER,
-    WCA_REQUEST_USER_UUID_HEADER,
     WCABaseCompletionsPipeline,
     WCABaseContentMatchPipeline,
     WCABaseMetaData,
@@ -115,13 +114,12 @@ def __init__(self, config: WCAOnPremConfiguration):
         # User may provide an override value if the setting is not defined.
 
     def get_request_headers(
-        self, api_key: str, identifier: Optional[str], lightspeed_user_uuid: Optional[str] = None
+        self, api_key: str, identifier: Optional[str]
     ) -> dict[str, Optional[str]]:
         base_headers = self._get_base_headers(api_key)
         return {
             **base_headers,
             WCA_REQUEST_ID_HEADER: str(identifier) if identifier else None,
-            WCA_REQUEST_USER_UUID_HEADER: lightspeed_user_uuid if lightspeed_user_uuid else None,
         }
 
     def _get_base_headers(self, api_key: str) -> dict[str, str]:

ansible_ai_connect/ai/api/model_pipelines/wca/pipelines_saas.py

@@ -57,7 +57,6 @@
 from ansible_ai_connect.ai.api.model_pipelines.wca.pipelines_base import (
     MODEL_MESH_HEALTH_CHECK_TOKENS,
     WCA_REQUEST_ID_HEADER,
-    WCA_REQUEST_USER_UUID_HEADER,
     WCABaseCompletionsPipeline,
     WCABaseContentMatchPipeline,
     WCABaseMetaData,
@@ -230,13 +229,12 @@ def __init__(self, config: WCASaaSConfiguration):
         super().__init__(config=config)
 
     def get_request_headers(
-        self, api_key: str, identifier: Optional[str], lightspeed_user_uuid: Optional[str] = None
+        self, api_key: str, identifier: Optional[str]
     ) -> dict[str, Optional[str]]:
         base_headers = self._get_base_headers(api_key)
         return {
             **base_headers,
             WCA_REQUEST_ID_HEADER: str(identifier) if identifier else None,
-            WCA_REQUEST_USER_UUID_HEADER: lightspeed_user_uuid if lightspeed_user_uuid else None,
         }
 
     def _get_base_headers(self, api_key: str) -> dict[str, str]: