test: Add test case for output tensor size (#8639)

Author: mattwittwer

qa/L0_parameters/class_count_test.py

@@ -1,6 +1,6 @@
 #!/usr/bin/env python3
 
-# Copyright 2025, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# Copyright 2025-2026, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
 #
 # Redistribution and use in source and binary forms, with or without
 # modification, are permitted provided that the following conditions
@@ -100,6 +100,19 @@ def test_classificattion_unsupported_data_type(self):
             str(e.exception),
         )
 
+    def test_classification_output_tensor_too_large(self):
+        max_elements = 1_000_000
+        shape = (1, max_elements + 1)
+        dtype = "FP32"
+        model_name = "identity_fp32"
+        input_data = np.ones(shape, dtype=np.float32)
+
+        inputs, outputs = self._prepare_io(input_data, dtype)
+        with self.assertRaises(InferenceServerException) as e:
+            self.client.infer(model_name=model_name, inputs=inputs, outputs=outputs)
+
+        self.assertIn("classification output tensor too large", str(e.exception))
+
 
 if __name__ == "__main__":
     unittest.main()

qa/L0_parameters/test.sh

@@ -1,5 +1,5 @@
 #!/bin/bash
-# Copyright 2023-2025, NVIDIA CORPORATION. All rights reserved.
+# Copyright 2023-2026, NVIDIA CORPORATION. All rights reserved.
 #
 # Redistribution and use in source and binary forms, with or without
 # modification, are permitted provided that the following conditions
@@ -136,7 +136,7 @@ for client_type in http grpc; do
       echo -e "\n***\n*** Test Failed - class_count_${client_type}_test_client\n***"
       RET=1
   else
-      check_test_results $TEST_RESULT_FILE 2
+      check_test_results $TEST_RESULT_FILE 3
       if [ $? -ne 0 ]; then
           cat $TEST_RESULT_FILE
           echo -e "\n***\n*** Test Result Verification Failed - class_count_${client_type}_test_client\n***"