[TEST]add a qwen3-30b acc case with mooncake mempool

[jiangyunfan1]

What this PR does / why we need it?

This PR adds a case of qwen3-30b w8a8 with mooncake mempool, we need to test it regual

Does this PR introduce any user-facing change?

No

How was this patch tested?

by running the test

• vLLM version: v0.14.1
• vLLM main: vllm-project/vllm@d682094

[github-actions]

👋 Hi! Thank you for contributing to the vLLM Ascend project. The following points will speed up your PR merge:‌‌

• A PR should do only one thing, smaller PRs enable faster reviews.
• Every PR should include unit tests and end-to-end tests ‌to ensure it works and is not broken by other future PRs.
• Write the commit message by fulfilling the PR description to help reviewer and future developers understand.

If CI fails, you can run linting and testing checks locally according Contributing and Testing.

[gemini-code-assist]

Code Review

This pull request adds a new end-to-end test for the Qwen3-30B-A3B-W8A8 model. My review identifies a potential issue with test robustness due to the use of a hardcoded file name for configuration. I've provided a suggestion to use pytest's tmp_path fixture to ensure tests are isolated and do not leave artifacts, which is a standard practice for writing robust tests.

[gemini-code-assist]

Creating a file with a hardcoded name (mooncake.json) in the current working directory is not a good practice for tests. It can lead to race conditions if tests are run in parallel and can leave artifacts if a test fails. It's better to use the tmp_path fixture provided by pytest to create temporary files in a managed way. This suggestion refactors the code to use tmp_path for creating the mooncake.json configuration file, which will make the test more robust and isolated.

Suggested change

	async def test_models(model: str, tp_size: int) -> None:
	port = get_open_port()
	mooncake_port = get_open_port()
	mooncake_metrics_port = get_open_port()
	mooncake_json["master_server_address"] = f"127.0.0.1:{mooncake_port}"
	with open("mooncake.json", "w") as f:
	json.dump(mooncake_json, f)
	env_dict = {
	"PYTHONHASHSEED": "0",
	"ASCEND_CONNECT_TIMEOUT": "10000",
	"ASCEND_TRANSFER_TIMEOUT": "10000",
	"ASCEND_BUFFER_POOL": "4:8",
	"VLLM_USE_V1": "1",
	"OMP_PROC_BIND": "false",
	"HCCL_OP_EXPANSION_MODE": "AIV",
	"HCCL_BUFFSIZE": "1024",
	"OMP_NUM_THREADS": "1",
	"PYTORCH_NPU_ALLOC_CONF": "expandable_segments:True",
	"VLLM_ASCEND_ENABLE_NZ": "2",
	"MOONCAKE_CONFIG_PATH": "mooncake.json"
	async def test_models(model: str, tp_size: int, tmp_path) -> None:
	port = get_open_port()
	mooncake_port = get_open_port()
	mooncake_metrics_port = get_open_port()
	mooncake_json["master_server_address"] = f"127.0.0.1:{mooncake_port}"
	mooncake_config_path = tmp_path / "mooncake.json"
	mooncake_config_path.write_text(json.dumps(mooncake_json))
	env_dict = {
	"PYTHONHASHSEED": "0",
	"ASCEND_CONNECT_TIMEOUT": "10000",
	"ASCEND_TRANSFER_TIMEOUT": "10000",
	"ASCEND_BUFFER_POOL": "4:8",
	"VLLM_USE_V1": "1",
	"OMP_PROC_BIND": "false",
	"HCCL_OP_EXPANSION_MODE": "AIV",
	"HCCL_BUFFSIZE": "1024",
	"OMP_NUM_THREADS": "1",
	"PYTORCH_NPU_ALLOC_CONF": "expandable_segments:True",
	"VLLM_ASCEND_ENABLE_NZ": "2",
	"MOONCAKE_CONFIG_PATH": str(mooncake_config_path)

[github-actions]

This pull request has conflicts, please resolve those before we can evaluate the pull request.