PR #39777: add prot available check when create server before

• Repository: vllm-project/vllm
• Author: lengrongfu
• State: open | merged: False
• Link: https://github.com/vllm-project/vllm/pull/39777

Description (problem / solution / changelog)

Purpose

Fixs: https://github.com/vllm-project/vllm/issues/39762

Test Plan

When can first step, start a vllm instance, can set --api-server-count=4, after, agent start a new vllm instance, when not set --port we can look a RuntimeError.

(APIServer pid=69077) RuntimeError: Port 8000 on 0.0.0.0 is already in use by another process. Please stop the conflicting service or use a different port (--port).

1. first step, start first a vllm instance

$ vllm serve /home/jovyan/qwen3-8b/ --api-server-count 4

start after, we can look this api server have four process. <img width="1312" height="256" alt="image" src="https://github.com/user-attachments/assets/e54a65e9-26f1-4364-a9d1-a7e0e6b68614" />

1. second step, start second vllm instance

$ vllm serve /home/jovyan/qwen3-8b/ --api-server-count 4

wen can't start success, can get RuntimeError, Port 8000 on 0.0.0.0 is already in use by another process. Please stop the conflicting service or use a different port (--port)

Test Result

This resolves situations where the port is already occupied by another process prior to startup.

• The purpose of the PR, such as "Fix some issue (link existing issues this PR will resolve)".
• The test plan, such as providing test command.
• The test results, such as pasting the results comparison before and after, or e2e results
• (Optional) The necessary documentation update, such as updating supported_models.md and examples for a new model.
• (Optional) Release notes update. If your change is user facing, please update the release notes draft in the Google Doc.

Changed files

• vllm/entrypoints/openai/api_server.py (modified, +18/-0)