[dpc]: unify DP controller load balancing and simplify dispatch logic

[Ratish1]

Motivation

This PR refactors the Data Parallel Controller (DPC) to align with the new load balancing guidelines and simplify the logic. #16080

Modifications

• Rename shortest_queue to total_requests and minimum_tokens to total_tokens.
• Simplify DPBudget to use a greedy dispatching strategy based on piggyback reports.
• Remove unnecessary ahead-of-time planning variables.
• Update documentation for Ascend NPU and server arguments.

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.
• Follow the SGLang code style guidance.

Review Process

1. Ping Merge Oncalls to start the PR flow. See the PR Merge Process.
2. Get approvals from CODEOWNERS and other reviewers.
3. Trigger CI tests with comments (/tag-run-ci-label, /rerun-failed-ci, /tag-and-rerun-ci) or contact authorized users to do so.
4. After green CI and required approvals, ask Merge Oncalls to merge.

[gemini-code-assist]

Summary of Changes

Hello @Ratish1, 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 streamlines the Data Parallel Controller (DPC) by modernizing its load balancing mechanisms. It renames existing load balancing strategies to better represent their functionality and drastically simplifies the internal dispatching logic within the DPBudget class. This refactoring aims to improve clarity and efficiency in how requests are distributed across data parallel workers, along with updating all relevant documentation.

Highlights

• Load Balancing Method Renaming: The shortest_queue and minimum_tokens load balancing methods have been renamed to total_requests and total_tokens respectively, reflecting their underlying metrics.
• Simplified DPBudget Logic: The DPBudget class has been significantly simplified, removing complex queue management and ahead-of-time planning variables, and now employs a greedy dispatching strategy based on real-time request and token counts.
• Documentation Updates: The server arguments and Ascend NPU support documentation have been updated to reflect the new load balancing method names.

🧠 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 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 by creating a comment using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands on the current page.

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 pull request 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 successfully refactors the Data Parallel Controller, simplifying the load balancing logic and improving clarity by renaming methods. The new greedy dispatch strategy in DPBudget is a significant improvement over the previous complex queue-based implementation. The documentation updates are also consistent with the code changes.

I have identified a potential issue in the dispatch logic for the TOTAL_TOKENS strategy that could cause load imbalance, and I've provided a suggestion to address it. Additionally, there's a minor suggestion to improve consistency in the command-line argument definitions. Overall, this is a solid refactoring that enhances maintainability.