Fix e2e data parallel test: add resource release code (#1881)

### What this PR does / why we need it?
Fix e2e data parallel test: add resource release code and give more time
to engine to pause their processing loops before exiting.

### Does this PR introduce _any_ user-facing change?
No

- vLLM version: v0.9.2
- vLLM main:
5895afd780

Signed-off-by: leo-pony <nengjunma@outlook.com>

examples/offline_data_parallel.py

@@ -56,14 +56,19 @@ Multi-node:
 
 import os
 from time import sleep
+import contextlib
+import gc
+
+import torch
 
 from vllm import LLM, SamplingParams
 from vllm.utils import get_open_port
+from vllm.distributed.parallel_state import (  # noqa E402
+    destroy_distributed_environment, destroy_model_parallel)
 
 os.environ["VLLM_USE_MODELSCOPE"] = "True"
 os.environ["VLLM_WORKER_MULTIPROC_METHOD"] = "spawn"
 
-
 def parse_args():
     import argparse
 
@@ -110,6 +115,15 @@ def parse_args():
     return parser.parse_args()
 
 
+def cleanup_env_and_memory():
+    destroy_model_parallel()
+    destroy_distributed_environment()
+    with contextlib.suppress(AssertionError):
+        torch.distributed.destroy_process_group()
+    gc.collect()
+    torch.npu.empty_cache()
+    torch.npu.reset_peak_memory_stats()
+
 def main(
     model,
     dp_size,
@@ -185,8 +199,9 @@ def main(
               f"Generated text: {generated_text!r}")
 
     # Give engines time to pause their processing loops before exiting.
-    sleep(1)
+    sleep(5)
-
+    del llm
+    cleanup_env_and_memory()
 
 if __name__ == "__main__":
     args = parse_args()

tests/e2e/multicard/test_data_parallel.py

@@ -30,7 +30,6 @@ import pytest
 MODELS = ["Qwen/Qwen2.5-0.5B-Instruct"]
 
 
-@pytest.mark.skipif(True, reason="TODO: fix dp timeout error in ci")
 @pytest.mark.parametrize("model", MODELS)
 @pytest.mark.parametrize("max_tokens", [32])
 @patch.dict(os.environ, {"ASCEND_RT_VISIBLE_DEVICES": "0,1"})