bugfix[schedule]: Excessive preemption occurs when preempting running requests to schedule new prefill requests.#12494
Conversation
Signed-off-by: CLFutureX <chenyongqyl@163.com>
Summary of ChangesHello @CLFutureX, 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 an issue of excessive preemption in the scheduling policy by refining the logic within the Highlights
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
This pull request introduces several optimizations to the request preemption logic in the scheduler. The main fix prevents excessive preemption by exiting the loop as soon as enough token space has been freed for the new request. Additionally, the code is optimized by filtering out non-preemptible requests before sorting, and by breaking the preemption loop early if a request does not meet the priority threshold, since subsequent requests in the sorted list also won't. These changes are well-implemented and should improve scheduling efficiency and prevent unnecessary preemptions.
|
@harrisonlimh Hey,PTAL |
|
@hnyls2002 @merrymercy @xiezhq-hermann @Ying1123 hey, PTAL , thanks |
| Returns True if preemption was committed, and the new request can be scheduled. | ||
| """ | ||
| # Iterate running requests to find preemptible requests | ||
| valid_running_reqs = ( |
There was a problem hiding this comment.
nit: can we filter for priority here as well?
|
LGTM! Thank you! |
|
@hnyls2002 can you take a look too? |
|
@hnyls2002 ping |
|
I will cover this bug fix test cases in https://github.com/sgl-project/sglang/pull/13576/files test_prefill_adder.py |
… requests to schedule new prefill requests. (sgl-project#12494) Signed-off-by: CLFutureX <chenyongqyl@163.com>
… requests to schedule new prefill requests. (sgl-project#12494) Signed-off-by: CLFutureX <chenyongqyl@163.com>
issue
Problem
Excessive preemption occurs when preempting running requests to schedule new prefill requests.
Affected Path
python/sglang/srt/managers/schedule_policy.pyRelevant Code Snippet
Solution
Exit the loop and stop preemption once the current prefill request's resource requirements are met (i.e., when
min_tokens_to_remove <= 0).Additional Adjustments