-
-
Notifications
You must be signed in to change notification settings - Fork 10.6k
[Misc] Getting and passing ray runtime_env to workers #22040
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
👋 Hi! Thank you for contributing to the vLLM project. 💬 Join our developer Slack at https://slack.vllm.ai to discuss your PR in #pr-reviews, coordinate on features in #feat- channels, or join special interest groups in #sig- channels. Just a reminder: PRs would not trigger full CI run by default. Instead, it would only run Once the PR is approved and ready to go, your PR reviewer(s) can run CI to test the changes comprehensively before merging. To run CI, PR reviewers can either: Add 🚀 |
There was a problem hiding this 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 introduces a mechanism to capture the Ray runtime_env
when creating an engine configuration. This is a valuable feature for distributed deployments on Ray. The implementation correctly captures the runtime_env
and stores it in the ParallelConfig
. However, my review found a critical issue where this captured runtime_env
is not actually used to configure the Ray workers, making the feature incomplete. The workers' runtime environment will not be affected by this change as it stands.
4502c08
to
20f243d
Compare
Signed-off-by: Rui Qiao <[email protected]>
Signed-off-by: Rui Qiao <[email protected]>
Signed-off-by: Rui Qiao <[email protected]>
…2040) Signed-off-by: Rui Qiao <[email protected]>
…2040) Signed-off-by: Rui Qiao <[email protected]> Signed-off-by: Jinzhen Lin <[email protected]>
…2040) Signed-off-by: Rui Qiao <[email protected]> Signed-off-by: Noam Gat <[email protected]>
…2040) Signed-off-by: Rui Qiao <[email protected]> Signed-off-by: Paul Pak <[email protected]>
…2040) Signed-off-by: Rui Qiao <[email protected]> Signed-off-by: Diego-Castan <[email protected]>
…2040) Signed-off-by: Rui Qiao <[email protected]>
…2040) Signed-off-by: Rui Qiao <[email protected]>
# as opposed to is_in_ray_actor(). | ||
import ray | ||
ray_runtime_env = ray.get_runtime_context().runtime_env | ||
logger.info("Using ray runtime env: %s", ray_runtime_env) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Please don't print things like envs to logs, they often can contain secrets.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks, make sense. I created a PR to address this. #26302
Essential Elements of an Effective PR Description Checklist
supported_models.md
andexamples
for a new model.Purpose
Currently vLLM does not pass down Ray runtime_env to the workers. This PR fixes it.
Fixes ray-project/ray#54866
Co-authored with @eric-higgins-ai from #21143
Test Plan
Added new unit test.
Manual test of #54866
Verified flashinfer dep is passed down and taking effect:
Test Result
Passed
(Optional) Documentation Update