diff --git a/qa/L0_model_queue/model_queue_test.py b/qa/L0_model_queue/model_queue_test.py index 14d2349c8c..e7be471f79 100755 --- a/qa/L0_model_queue/model_queue_test.py +++ b/qa/L0_model_queue/model_queue_test.py @@ -140,7 +140,7 @@ def test_max_queue_size(self): for trial in self.trials_: preceding_thread = threading.Thread( target=self.check_response, - args=(8, dtype, shapes, 0, 0, (1999, 1000)), + args=(8, dtype, shapes, 0, 0, (5999, 1000)), ) threads = [] for i in range(10): @@ -160,8 +160,10 @@ def test_max_queue_size(self): for t in threads: t.join() - # Expect at most two exception with exceeding max queue size error - for i in range(2): + # Expect exactly two exception with exceeding max queue size error + expected_exceeded_count = 2 + exceeded_count = 0 + for i in range(expected_exceeded_count): try: self.check_deferred_exception() except InferenceServerException as ex: @@ -171,6 +173,14 @@ def test_max_queue_size(self): ex ), ) + exceeded_count = exceeded_count + 1 + self.assertEqual( + exceeded_count, + expected_exceeded_count, + "expected {} requests to fail with exceeded max queue size error, got {}".format( + expected_exceeded_count, exceeded_count + ), + ) try: self.check_deferred_exception() except InferenceServerException as ex: diff --git a/qa/L0_model_queue/test.sh b/qa/L0_model_queue/test.sh index 667f902db0..577b7b7fc2 100755 --- a/qa/L0_model_queue/test.sh +++ b/qa/L0_model_queue/test.sh @@ -82,7 +82,7 @@ rm -fr models && mkdir models && \ echo " }" >> config.pbtxt && \ echo "}" >> config.pbtxt && \ echo "parameters [" >> config.pbtxt && \ - echo "{ key: \"execute_delay_ms\"; value: { string_value: \"1000\" }}" >> config.pbtxt && \ + echo "{ key: \"execute_delay_ms\"; value: { string_value: \"5000\" }}" >> config.pbtxt && \ echo "]" >> config.pbtxt) TEST_CASE=test_max_queue_size