Fix: /flush_cache API hangs when using --skip-server-warmup.

[solrex]

Motivation

The /flush_cache API is currently no-return when the server is launched with the --skip-server-warmup argument, causing the client to hang.

python3 -m sglang.launch_server --model-path ANY_MODEL --disable-cuda-graph --skip-server-warmup
curl http://localhost:30000/flush_cache 

Modifications

• Add self.auto_create_handle_loop() when /flush_cache is called.

Accuracy Tests

Benchmarking and Profiling

Checklist

• Format your code according to the Format code with pre-commit.
• Add unit tests according to the Run and add unit tests.
• Update documentation according to Write documentations.
• Provide accuracy and speed benchmark results according to Test the accuracy and Benchmark the speed.

[gemini-code-assist]

Summary of Changes

Hello @solrex, I'm Gemini Code Assist¹! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

This pull request addresses a critical issue where the "/flush_cache" API would hang indefinitely when the server was started with the "--skip-server-warmup" flag. The fix ensures that the internal handle loop, which is crucial for the API's operation, is properly initialized regardless of the server's warm-up status. This resolves the unresponsiveness and allows the cache flushing operation to complete successfully.

Highlights

• API Fix: The "/flush_cache" API no longer hangs when the server is launched with the "--skip-server-warmup" argument, ensuring the API behaves as expected.
• Server Warmup: Ensures that the necessary internal handle loop is initialized even when the server's warm-up process is skipped, preventing the API from becoming unresponsive.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point in your pull request via creating an issue comment (i.e. comment on the pull request page) using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands.

Feature	Command	Description
Code Review	/gemini review	Performs a code review for the current pull request in its current state.
Pull Request Summary	/gemini summary	Provides a summary of the current pull request in its current state.
Comment	@gemini-code-assist	Responds in comments when explicitly tagged, both in issue comments and review comments.
Help	/gemini help	Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Footnotes

1. Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

[gemini-code-assist]

Code Review

This pull request addresses a bug where the /flush_cache API would hang if the server was started with --skip-server-warmup. The root cause was that the event loop responsible for handling responses was not being started in this scenario. The fix, which involves adding a call to self.auto_create_handle_loop() at the beginning of the flush_cache method, is correct and effectively resolves the issue. This change aligns flush_cache with the pattern used by other asynchronous methods in the TokenizerManager class that await responses. The implementation is clean and I have no further comments on the change itself. As a side note, while reviewing the code, I noticed that the get_internal_state and set_internal_state methods might be susceptible to a similar hanging issue, as they also appear to await a response without ensuring the handle_loop is running. It might be worth investigating these methods as well in a separate change.