diff --git a/release/llm_tests/serve/test_llm_serve_multi_node_integration.py b/release/llm_tests/serve/test_llm_serve_multi_node_integration.py
index 4e4e0dfc2b32..07ff92eb4f6b 100644
--- a/release/llm_tests/serve/test_llm_serve_multi_node_integration.py
+++ b/release/llm_tests/serve/test_llm_serve_multi_node_integration.py
@@ -3,13 +3,11 @@
 import ray
 from ray import serve
 from ray.serve.llm import (
+    build_dp_deployment,
+    build_openai_app,
     LLMConfig,
     LLMServingArgs,
     ModelLoadingConfig,
-    build_openai_app,
-)
-from ray.llm._internal.serve.serving_patterns.data_parallel.dp_server import (
-    build_dp_deployment,
 )
 
 
@@ -100,6 +98,9 @@ def test_llm_serve_data_parallelism():
             max_num_batched_tokens=256,
             enforce_eager=True,
         ),
+        experimental_configs=dict(
+            dp_size_per_node=2,
+        ),
         placement_group_config=placement_group_config,
         runtime_env={"env_vars": {"VLLM_DISABLE_COMPILE_CACHE": "1"}},
     )