feat: add forward timeout#17831
Conversation
Summary of ChangesHello @zhooooong, 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 enhances the SGLang server's robustness by integrating a configurable timeout mechanism for requests during their forward processing. This feature prevents requests from consuming excessive resources or hanging indefinitely, thereby improving the overall stability and responsiveness of the server by gracefully handling long-running or stuck operations. Highlights
🧠 New Feature in Public Preview: You can now enable Memory to help Gemini Code Assist learn from your team's feedback. This makes future code reviews more consistent and personalized to your project's style. Click here to enable Memory in your admin console. Using Gemini Code AssistThe 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 by creating a comment using either
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 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
|
There was a problem hiding this comment.
Code Review
The pull request introduces a new environment variable SGLANG_FORWARD_TIMEOUT_MS to control the timeout for requests in the forward batch. The implementation correctly integrates this timeout into both prefill and decode stages of the scheduler, aborting requests that exceed the configured duration. A dedicated unit test has been added to verify the functionality, ensuring that requests time out as expected. The documentation has also been updated to reflect the new environment variable. The changes are well-implemented and thoroughly tested.
|
/tag-and-rerun-ci |
Co-authored-by: qiuxuan.lzw <qiuxuan.lzw@alibaba-inc.com>
Co-authored-by: qiuxuan.lzw <qiuxuan.lzw@alibaba-inc.com>
Co-authored-by: qiuxuan.lzw <qiuxuan.lzw@alibaba-inc.com>
Co-authored-by: qiuxuan.lzw <qiuxuan.lzw@alibaba-inc.com>
Motivation
Introduce a timeout for requests in the forward batch.
Modifications
Introduce a new environment variable
SGLANG_FORWARD_TIMEOUT_MS. In the scheduler, abort requests that have exceeded the forward timeout before the next decode step.Accuracy Tests
server cmd:
Then, send a few requests casually. Subsequently, the requests encountered the error, as expected:
{"object":"error","message":"Forward timeout.","type":"HTTPStatus.SERVICE_UNAVAILABLE","param":null,"code":503}Benchmarking and Profiling
Checklist
Review Process
/tag-run-ci-label,/rerun-failed-ci,/tag-and-rerun-ci