Merge branch '2.25' into backport-operator-health-marker-2.25

Author: dbasunag

docs/GETTING_STARTED.md

@@ -122,11 +122,6 @@ For example, to check only `Serveless` and `Service Mesh` operators, pass `--tc=
 To run tests with admin client only, pass `--tc=use_unprivileged_client:False` to pytest.
 
 
-### jira integration
-To skip running tests which have open bugs, [pytest_jira](https://github.com/rhevm-qe-automation/pytest_jira) plugin is used.
-To run tests with jira integration, you need to set `PYTEST_JIRA_URL` and `PYTEST_JIRA_TOKEN` environment variables.
-To make a test with jira marker, add: `@pytest.mark.jira(jira_id="RHOAIENG-0000", run=False)` to the test.
-
 
 ### Running containerized tests
 Save kubeconfig file to a local directory, for example: `$HOME/kubeconfig`

pyproject.toml

@@ -57,7 +57,6 @@ dependencies = [
     "grpcio-reflection",
     "portforward>=0.7.1",
     "pytest-testconfig>=0.2.0",
-    "pytest-jira>=0.3.21",
     "pygithub>=2.5.0",
     "timeout-sampler>=1.0.6",
     "shortuuid>=1.0.13",
@@ -66,7 +65,6 @@ dependencies = [
     "semver>=3.0.4",
     "sqlalchemy>=2.0.40",
     "pytest-order>=1.3.0",
-    "marshmallow==3.26.1,<4", # this version is needed for pytest-jira
     "pytest-html>=4.1.1",
     "fire",
     "llama_stack_client==0.2.23",

pytest.ini

@@ -5,7 +5,6 @@ testpaths = tests
 markers =
     # General
     polarion: Store polarion test ID
-    jira: Store jira bug ID
     skip_on_disconnected: Mark tests that can only be run in deployments with Internet access i.e. not on disconnected clusters.
 
     # CI
@@ -59,4 +58,3 @@ addopts =
     --show-progress
     --tc-file=tests/global_config.py
     --tc-format=python
-    --jira

tests/model_serving/model_runtime/triton/basic_model_deployment/conftest.py

@@ -157,7 +157,7 @@ def triton_serving_runtime(
     admin_client: DynamicClient,
     model_namespace: Namespace,
     protocol: str,
-    supported_accelerator_type: str,
+    supported_accelerator_type: str | None,
 ) -> Generator[ServingRuntime, None, None]:
     template_name = get_template_name(protocol=protocol, accelerator_type=supported_accelerator_type)
     with ServingRuntimeFromTemplate(
@@ -178,7 +178,7 @@ def triton_inference_service(
     triton_serving_runtime: ServingRuntime,
     s3_models_storage_uri: str,
     triton_model_service_account: ServiceAccount,
-    supported_accelerator_type: str,
+    supported_accelerator_type: str | None,
 ) -> Generator[InferenceService, Any, Any]:
     params = request.param
     model_format = params.get(

tests/model_serving/model_runtime/triton/basic_model_deployment/utils.py

@@ -155,15 +155,20 @@ def validate_inference_request(
     assert response == response_snapshot, f"Output mismatch: {response} != {response_snapshot}"
 
 
-def get_gpu_identifier(accelerator_type: str) -> str:
+def get_gpu_identifier(accelerator_type: str | None) -> str:
+    if accelerator_type is None:
+        return Labels.Nvidia.NVIDIA_COM_GPU
     return ACCELERATOR_IDENTIFIER.get(accelerator_type.lower(), Labels.Nvidia.NVIDIA_COM_GPU)
 
 
-def get_template_name(protocol: str, accelerator_type: str) -> str:
+def get_template_name(protocol: str, accelerator_type: str | None) -> str:
     """
     Returns template name based on protocol and accelerator type.
     Falls back to default TRITON_REST if not found.
+    If accelerator_type is None or empty, defaults to "nvidia".
     """
+    if not (accelerator_type and accelerator_type.strip()):
+        accelerator_type = "nvidia"
     key = f"{protocol.lower()}_{accelerator_type.lower()}"
     return TEMPLATE_MAP.get(key, RuntimeTemplates.TRITON_REST)
 

tests/model_serving/model_server/authentication/test_kserve_token_authentication_raw.py

@@ -53,7 +53,6 @@ def test_disabled_raw_model_authentication(self, patched_remove_raw_authenticati
         )
 
     @pytest.mark.sanity
-    @pytest.mark.jira("RHOAIENG-19275", run=False)
     def test_raw_disable_enable_authentication_no_pod_rollout(self, http_s3_caikit_raw_inference_service):
         """Verify no pod rollout when disabling and enabling authentication"""
         pod = get_pods_by_isvc_label(

tests/model_serving/model_server/components/test_custom_resources.py

@@ -34,7 +34,6 @@ def wait_for_isvc_model_status(isvc: InferenceService, target_model_state: str,
         raise
 
 
-@pytest.mark.jira("RHOAIENG-10765")
 @pytest.mark.parametrize(
     "unprivileged_model_namespace, serving_runtime_from_template, invalid_s3_models_inference_service",
     [

tests/model_serving/model_server/model_car/test_oci_image.py

@@ -43,7 +43,6 @@
 )
 class TestKserveModelCar:
     @pytest.mark.smoke
-    @pytest.mark.jira("RHOAIENG-13465")
     def test_model_car_no_restarts(self, model_car_inference_service):
         """Verify that model pod doesn't restart"""
         pod = get_pods_by_isvc_label(
@@ -57,7 +56,6 @@ def test_model_car_no_restarts(self, model_car_inference_service):
 
     @pytest.mark.smoke
     @pytest.mark.ocp_interop
-    @pytest.mark.jira("RHOAIENG-12306")
     def test_model_car_using_rest(self, model_car_inference_service):
         """Verify model query with token using REST"""
         verify_inference_response(

tests/model_serving/model_server/ovms/kserve/test_onnx_serverless.py

@@ -33,7 +33,6 @@
 )
 class TestONNXServerless:
     @pytest.mark.smoke
-    @pytest.mark.jira("RHOAIENG-9045")
     def test_serverless_onnx_rest_inference(self, ovms_kserve_inference_service):
         """Verify that kserve Serverless ONNX model can be queried using REST"""
         verify_inference_response(

tests/model_serving/model_server/raw_deployment/test_bge_large_eng_caikit.py

@@ -14,7 +14,6 @@
 pytestmark = [pytest.mark.rawdeployment, pytest.mark.usefixtures("valid_aws_config")]
 
 
-@pytest.mark.jira("RHOAIENG-11749")
 @pytest.mark.parametrize(
     "unprivileged_model_namespace, serving_runtime_from_template, s3_models_inference_service",
     [