Skip to content

Conversation

@jiangxiaobin96
Copy link
Contributor

Unify code for selecting minimum value code which is the same with least_util and least_gpu_cache and avoid two for loop.

Copy link
Contributor

@gemini-code-assist gemini-code-assist bot left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Summary of Changes

Hello @jiangxiaobin96, 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!

I've refactored the code responsible for selecting the minimum value in the least_request algorithm. The primary goal was to unify the logic that was previously duplicated across similar functions and, more importantly, to eliminate a redundant loop, making the code more efficient.

Highlights

  • Code Unification & Optimization: I've optimized the selectTargetPodWithLeastRequestCount function in pkg/plugins/gateway/algorithms/least_request.go by consolidating two separate loops into a single, more efficient iteration. This change streamlines the process of identifying pods with the minimum request count.
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 or fill out our survey to provide feedback.

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.

Copy link
Contributor

@gemini-code-assist gemini-code-assist bot left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Code Review

The pull request aims to unify the code for selecting the minimum value by refactoring two loops into a single one. While this is a good optimization for performance, the new implementation contains a logical flaw that will cause incorrect behavior. The condition to find the minimum value should be a strict less-than comparison, not less-than-or-equal, to correctly handle cases where multiple pods have the same minimum request count. I've provided a critical review comment with a suggested fix to address this issue.

Comment on lines 85 to 90
if totalReq <= minCount {
minCount = totalReq
}
}
for podname, totalReq := range podRequestCount {
if totalReq == minCount {
targetPods = []string{podname}
} else if totalReq == minCount {
targetPods = append(targetPods, podname)
}
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

critical

The logic for finding the pods with the minimum request count is flawed. The condition totalReq <= minCount will cause targetPods to be reset every time a pod with the same minimum count is found, effectively only keeping the last one encountered. Additionally, the else if totalReq == minCount block is unreachable because the first condition totalReq <= minCount already covers the equality case.

To fix this, you should use a strict less-than comparison (<) to identify a new minimum, and handle the equality case to append pods with the same minimum count.

Suggested change
if totalReq <= minCount {
minCount = totalReq
}
}
for podname, totalReq := range podRequestCount {
if totalReq == minCount {
targetPods = []string{podname}
} else if totalReq == minCount {
targetPods = append(targetPods, podname)
}
if totalReq < minCount {
minCount = totalReq
targetPods = []string{podname}
} else if totalReq == minCount {
targetPods = append(targetPods, podname)
}

@Jeffwan
Copy link
Collaborator

Jeffwan commented Aug 10, 2025

/cc @varungup90 @Xunzhuo

Copy link
Collaborator

@Jeffwan Jeffwan left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

seem this is not reviewed yet. the change is very straightforward and it looks good to me.

Signed-off-by: jiangxiaobin96 <[email protected]>
@Jeffwan Jeffwan mentioned this pull request Aug 17, 2025
@Jeffwan Jeffwan changed the title Unified minimum value selection code Refactor to use single loop for least request pod selection Aug 17, 2025
@Jeffwan
Copy link
Collaborator

Jeffwan commented Aug 17, 2025

@jiangxiaobin96 thanks for the improvement and apologies for the delay in reviewing. I changed the PR title a little bit to make it more self-explainable.

@Jeffwan Jeffwan merged commit 5b5088c into vllm-project:main Aug 17, 2025
14 checks passed
@jiangxiaobin96 jiangxiaobin96 deleted the least_request branch August 19, 2025 13:38
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants