Skip to content

Conversation

alikhabazian
Copy link

[Router/Bugfix] Fix dynamic config in vllm router

Fixed dynamic config functionality in vllm router to allow config changes without container downtime. The changes include:

  • Used app inside dynamic config
  • Added default None values for:
    • aliases
    • model_labels
    • model_types
    • static_backend_health_checks
    • prefill_model_labels
    • decode_model_labels

Testing:

  • Manually verified config changes work without container restart

Signed-off-by: [Ali Khabazian] <[[email protected]]>

Copy link

@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 @alikhabazian, 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 delivers a crucial bugfix to the vLLM router, addressing issues that prevented dynamic configuration updates from taking effect without a full service restart. By correctly passing the application instance and making several constructor parameters optional with default values, the changes enable seamless, on-the-fly adjustments to the router's static backend configurations, significantly improving operational agility and reducing downtime.

Highlights

  • Dynamic Configuration Bugfix: Resolved a critical bug in the vLLM router's dynamic configuration, specifically for static service discovery, enabling configuration changes to be applied without requiring a container restart.
  • Application Instance Passing: Ensured that the self.app instance is correctly passed to the reconfigure_service_discovery function when setting up static service discovery, which was a missing dependency for dynamic updates.
  • Constructor Parameter Flexibility: Updated the StaticServiceDiscovery constructor to include default None or False values for several optional parameters (e.g., aliases, model_labels, static_backend_health_checks). This change makes the constructor more robust and allows for partial configuration updates during dynamic reconfigurations.
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.

Copy link

@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

This pull request fixes dynamic configuration reloading. The changes correctly pass the app object during static service discovery reconfiguration and add default values to StaticServiceDiscovery's initializer to prevent crashes when optional parameters are missing from the dynamic configuration. The review identifies that model aliases are not being passed during reconfiguration in dynamic_config.py, and the type hint for the aliases parameter is incorrect in service_discovery.py.

@@ -139,6 +139,7 @@ def reconfigure_service_discovery(self, config: DynamicRouterConfig):
ServiceDiscoveryType.STATIC,
urls=parse_static_urls(config.static_backends),
models=parse_comma_separated_args(config.static_models),
app=self.app

Choose a reason for hiding this comment

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

high

The config.static_aliases from DynamicRouterConfig is not being passed to reconfigure_service_discovery. This means that model aliases will not be updated during a dynamic reconfiguration, which is a bug.

Pass the parsed aliases to reconfigure_service_discovery.

Suggested change
app=self.app
app=self.app,
aliases=(
parse_static_aliases(config.static_aliases)
if config.static_aliases
else None
),

@zerofishnoodles zerofishnoodles mentioned this pull request Jul 23, 2025
3 tasks
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant