[start up hang] hang when I run 1st example from readme, please help

Nobody has claimed this yet.

hello guys,

I already install latest vllm and lmcache, I can run through vllm v0 on cpu offload example very well. but failed to run 1st example from lmcache-test
```
he-tests# python3 -c "import vllm; print(vllm.__name__)"
INFO 05-03 01:32:11 [__init__.py:239] Automatically detected platform cuda.
vllm
(lmvllm) root@salab-hpedl380g11-02:/mnt/nvme2n1/wayne/lmvllm/lmcache-tests# python3 main.py tests/tests.py -f test_lmcache_local_cpu -o outputs/
[2025-05-03 01:33:12,855] INFO: Importing testing script tests... (main.py:57:__main__)
[2025-05-03 01:33:14,927] INFO: Filtering functions with pattern: test_lmcache_local_cpu (main.py:69:__main__)
[2025-05-03 01:33:14,928] INFO: Found function: test_lmcache_local_cpu (main.py:78:__main__)
[2025-05-03 01:33:14,928] INFO: Collected 1 functions to execute. (main.py:82:__main__)
[2025-05-03 01:33:14,928] INFO: Executing function: test_lmcache_local_cpu (main.py:95:__main__)
[2025-05-03 01:33:14,928] INFO: Running experiment: {"qps": 0.125, "duration": 80, "context_length": 8192, "query_length": 16, "offset": 0} Usecase.DUMMY (driver.py:316:driver)
Launching Engine with Command : LMCACHE_CONFIG_FILE=configs/lmcache_local_cpu.yaml vllm serve mistralai/Mistral-7B-Instruct-v0.2 --port 8000 --gpu-memory-utilization 0.8 --tensor-parallel-size 1 --trust-remote-code --kv-transfer-config '{"kv_connector":"LMCacheConnector", "kv_role":"kv_both"}'
Launching Engine with Command : vllm serve mistralai/Mistral-7B-Instruct-v0.2 --port 8001 --gpu-memory-utilization 0.8 --tensor-parallel-size 1 --trust-remote-code
[2025-05-03 01:33:16,618] DEBUG: Waiting for the pattern 'Application startup complete.' in the files ['/tmp/root-8000-stdout.log', '/tmp/root-8000-stderr.log'] (bootstrapper.py:70:bootstrapper)
[2025-05-03 01:33:17,619] DEBUG: Waiting for the pattern 'Application startup complete.' in the files ['/tmp/root-8000-stdout.log', '/tmp/root-8000-stderr.log'] (bootstrapper.py:70:bootstrapper)
[2025-05-03 01:33:18,620] DEBUG: Waiting for the pattern 'Application startup complete.' in the files ['/tmp/root-8000-stdout.log', '/tmp/root-8000-stderr.log'] (bootstrapper.py:70:bootstrapper)
[2025-05-03 01:33:19,621] DEBUG: Waiting for the pattern 'Application startup complete.' in the files ['/tmp/root-8000-stdout.log', '/tmp/root-8000-stderr.log'] (bootstrapper.py:70:bootstrapper)
[2025-05-03 01:33:20,621] DEBUG: Waiting for the pattern 'Application startup complete.' in the files ['/tmp/root-8000-stdout.log', '/tmp/root-8000-stderr.log'] (bootstrapper.py:70:bootstrapper)
```

Start with the README's first example and trace test_lmcache_local_cpu from tests/tests.py through main.py and bootstrapper.py. Reproduce the command and inspect /tmp/root-8000-stdout.log and /tmp/root-8000-stderr.log while it waits for “Application startup complete.” Done means identifying and resolving the startup hang or documenting the required environment or configuration.

Written by the indexing model from the issue text.