[Router] Support load balancing in PD router

[zhouwfang]

Add three basic LB strategies in PD router.

The next PR will add such LB strategy in a command line argument.

BEFORE SUBMITTING, PLEASE READ THE CHECKLIST BELOW AND FILL IN THE DESCRIPTION ABOVE

---

• Make sure the code changes pass the pre-commit checks.
• Sign-off your commit by using -s when doing git commit
• Try to classify PRs for easy understanding of the type of changes, such as [Bugfix], [Feat], and [CI].

Detailed Checklist (Click to Expand)

Thank you for your contribution to production-stack! Before submitting the pull request, please ensure the PR meets the following criteria. This helps us maintain the code quality and improve the efficiency of the review process.

PR Title and Classification

Please try to classify PRs for easy understanding of the type of changes. The PR title is prefixed appropriately to indicate the type of change. Please use one of the following:

• [Bugfix] for bug fixes.
• [CI/Build] for build or continuous integration improvements.
• [Doc] for documentation fixes and improvements.
• [Feat] for new features in the cluster (e.g., autoscaling, disaggregated prefill, etc.).
• [Router] for changes to the vllm_router (e.g., routing algorithm, router observability, etc.).
• [Misc] for PRs that do not fit the above categories. Please use this sparingly.

Note: If the PR spans more than one category, please include all relevant prefixes.

Code Quality

The PR need to meet the following code quality standards:

• Pass all linter checks. Please use pre-commit to format your code. See README.md for installation.
• The code need to be well-documented to ensure future contributors can easily understand the code.
• Please include sufficient tests to ensure the change is stay correct and robust. This includes both unit tests and integration tests.

DCO and Signed-off-by

When contributing changes to this project, you must agree to the DCO. Commits must include a Signed-off-by: header which certifies agreement with the terms of the DCO.

Using -s with git commit will automatically add this header.

What to Expect for the Reviews

We aim to address all PRs in a timely manner. If no one reviews your PR within 5 days, please @-mention one of YuhanLiu11
, Shaoting-Feng or ApostaC.

[gemini-code-assist]

Summary of Changes

Hello @zhouwfang, 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 significantly enhances the PD router's capabilities by integrating robust load balancing mechanisms. It introduces configurable strategies, including Round Robin, Random, and QPS-based routing, allowing for more efficient distribution of prefill and decode requests across available endpoints. This change lays the groundwork for improved resource utilization and performance in the router.

Highlights

• New Load Balancing Strategies: Introduced a LoadBalancingStrategy enum (ROUND_ROBIN, RANDOM, QPS) to define the available load balancing methods for the PD router.
• Configurable Router Initialization: The DisaggregatedPrefillRouter now accepts a load_balancing_strategy parameter during initialization, defaulting to ROUND_ROBIN.
• Load Balancing Implementations: Added dedicated private methods (_select_endpoint_round_robin, _select_endpoint_random, _select_endpoint_qps) to encapsulate the logic for each load balancing algorithm.
• Dynamic Request Routing: The route_request method in DisaggregatedPrefillRouter has been updated to dynamically apply the configured load balancing strategy when selecting prefill or decode endpoints.

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 is currently in preview and 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 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. ↩

[gemini-code-assist]

Code Review

This PR introduces load balancing strategies to the disaggregated prefill router. The round-robin strategy has a race condition that could lead to incorrect load distribution in a concurrent environment. The request counter updates should be made thread-safe using locks.

[YuhanLiu11]

@zhouwfang This is an awesome PR! We should definitely support this in PD. Just wanted to make sure, does this PR assume XpYd? (more than one prefiller pods and decoder pods)

[zhouwfang]

@zhouwfang This is an awesome PR! We should definitely support this in PD. Just wanted to make sure, does this PR assume XpYd? (more than one prefiller pods and decoder pods)

@YuhanLiu11 This PR only adds loading balancing to the existing PD router, which is only able to route a request to a single P/D worker. Real XPYD is beyond the scope of this PR, and I'd be happy to work on it.

[YuhanLiu11]

@zhouwfang This is an awesome PR! We should definitely support this in PD. Just wanted to make sure, does this PR assume XpYd? (more than one prefiller pods and decoder pods)

@YuhanLiu11 This PR only adds loading balancing to the existing PD router, which is only able to route a request to a single P/D worker. Real XPYD is beyond the scope of this PR, and I'd be happy to work on it.

Ah got it! I'll review tomorrow on this PR. And yes XpYd deserves another PR and we could refer to https://github.com/LMCache/LMCache/tree/dev/examples/disagg_prefill/xpyd_experimental as an example design.