migrate to ovms (#904)

* remove kueue test from smoek 2.22

Signed-off-by: Milind Waykole <mwaykole@redhat.com>

* Fix tierdown deletion of project and migrate to ovms test from caikit

Signed-off-by: Milind Waykole <mwaykole@redhat.com>

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

* fix: remove unused Caikit fixtures after OVMS migration

* feat: convert remaining Caikit smoke tests to OVMS MNIST

- Convert test_model_metrics.py from CAIKIT_TGIS to OVMS ONNX
- Convert test_bge_large_eng_caikit.py to test_ovms_mnist_raw.py
- Update metrics queries from tgi_* to ovms_*
- Use ONNX_INFERENCE_CONFIG instead of CAIKIT_* configs

---------

Signed-off-by: Milind Waykole <mwaykole@redhat.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>

Author: mwaykole

tests/model_serving/model_server/metrics/test_model_metrics.py

@@ -7,13 +7,12 @@
 from utilities.constants import (
     KServeDeploymentType,
     ModelFormat,
-    ModelInferenceRuntime,
-    ModelStoragePath,
+    ModelVersion,
     Protocols,
     RuntimeTemplates,
 )
 from utilities.inference_utils import Inference
-from utilities.manifests.caikit_tgis import CAIKIT_TGIS_INFERENCE_CONFIG
+from utilities.manifests.onnx import ONNX_INFERENCE_CONFIG
 from utilities.monitoring import get_metrics_value, validate_metrics_field
 
 pytestmark = [
@@ -28,17 +27,18 @@
     "unprivileged_model_namespace, serving_runtime_from_template, s3_models_inference_service",
     [
         pytest.param(
-            {"name": "kserve-tgis-metrics"},
+            {"name": "kserve-ovms-metrics"},
             {
-                "name": f"{Protocols.HTTP}-{ModelInferenceRuntime.CAIKIT_TGIS_RUNTIME}",
-                "template-name": RuntimeTemplates.CAIKIT_TGIS_SERVING,
+                "name": f"{Protocols.HTTP}-{ModelFormat.ONNX}",
+                "template-name": RuntimeTemplates.OVMS_KSERVE,
                 "multi-model": False,
                 "enable-http": True,
             },
             {
-                "name": f"{Protocols.HTTP}-{ModelFormat.CAIKIT}",
+                "name": f"{Protocols.HTTP}-{ModelFormat.ONNX}",
                 "deployment-mode": KServeDeploymentType.SERVERLESS,
-                "model-dir": ModelStoragePath.FLAN_T5_SMALL_CAIKIT,
+                "model-dir": "test-dir",
+                "model-version": ModelVersion.OPSET13,
             },
         )
     ],
@@ -51,15 +51,14 @@ def test_model_metrics_num_success_requests(self, s3_models_inference_service, p
         """Verify number of successful model requests in OpenShift monitoring system (UserWorkloadMonitoring) metrics"""
         verify_inference_response(
             inference_service=s3_models_inference_service,
-            inference_config=CAIKIT_TGIS_INFERENCE_CONFIG,
-            inference_type=Inference.ALL_TOKENS,
+            inference_config=ONNX_INFERENCE_CONFIG,
+            inference_type=Inference.INFER,
             protocol=Protocols.HTTPS,
-            model_name=ModelFormat.CAIKIT,
             use_default_query=True,
         )
         validate_metrics_field(
             prometheus=prometheus,
-            metrics_query="tgi_request_success",
+            metrics_query="ovms_requests_success",
             expected_value="1",
         )
 
@@ -71,16 +70,15 @@ def test_model_metrics_num_total_requests(self, s3_models_inference_service, pro
 
         run_inference_multiple_times(
             isvc=s3_models_inference_service,
-            inference_config=CAIKIT_TGIS_INFERENCE_CONFIG,
-            inference_type=Inference.ALL_TOKENS,
+            inference_config=ONNX_INFERENCE_CONFIG,
+            inference_type=Inference.INFER,
             protocol=Protocols.HTTPS,
-            model_name=ModelFormat.CAIKIT,
             iterations=total_runs,
             run_in_parallel=True,
         )
         validate_metrics_field(
             prometheus=prometheus,
-            metrics_query="tgi_request_count",
+            metrics_query="ovms_requests_success",
             expected_value=str(total_runs + 1),
         )
 

tests/model_serving/model_server/raw_deployment/test_bge_large_eng_caikit.py

@@ -0,0 +1,49 @@
+import pytest
+
+from tests.model_serving.model_server.utils import verify_inference_response
+from utilities.constants import (
+    KServeDeploymentType,
+    ModelFormat,
+    ModelVersion,
+    Protocols,
+    RuntimeTemplates,
+)
+from utilities.inference_utils import Inference
+from utilities.manifests.onnx import ONNX_INFERENCE_CONFIG
+
+pytestmark = [pytest.mark.rawdeployment, pytest.mark.usefixtures("valid_aws_config")]
+
+
+@pytest.mark.smoke
+@pytest.mark.jira("RHOAIENG-11749")
+@pytest.mark.parametrize(
+    "unprivileged_model_namespace, serving_runtime_from_template, s3_models_inference_service",
+    [
+        pytest.param(
+            {"name": "raw-deployment-ovms-mnist"},
+            {
+                "name": f"{Protocols.HTTP}-{ModelFormat.ONNX}",
+                "template-name": RuntimeTemplates.OVMS_KSERVE,
+                "multi-model": False,
+                "enable-http": True,
+            },
+            {
+                "name": f"{Protocols.HTTP}-{ModelFormat.ONNX}",
+                "deployment-mode": KServeDeploymentType.RAW_DEPLOYMENT,
+                "model-dir": "test-dir",
+                "model-version": ModelVersion.OPSET13,
+            },
+        )
+    ],
+    indirect=True,
+)
+class TestOvmsMnistRaw:
+    def test_ovms_mnist_inference_raw_internal_route(self, s3_models_inference_service):
+        """Test OVMS MNIST model inference using internal route"""
+        verify_inference_response(
+            inference_service=s3_models_inference_service,
+            inference_config=ONNX_INFERENCE_CONFIG,
+            inference_type=Inference.INFER,
+            protocol=Protocols.HTTP,
+            use_default_query=True,
+        )