modified lws chart to support aws efa#212
Open
masoodfaisal wants to merge 2 commits intoawslabs:mainfrom
Open
Conversation
erezzarum
reviewed
Oct 31, 2025
| {{- if and ( eq .Values.inference.accelerator "gpu" ) ( eq .Values.inference.awsEfa true ) }} | ||
| - name: FI_PROVIDER | ||
| value: "efa" | ||
| - name: FI_EFA_USE_DEVICE_RDMA |
Contributor
There was a problem hiding this comment.
Why is this needed?
I believe TP/PP uses NCCL, in this case we should use the tuner plugin to adjust the correct parameters and not to force anything.
Please test with NCCL_TUNER_PLUGIN=ofi when using our DLCs
erezzarum
reviewed
Oct 31, 2025
| modelServer: | ||
| image: | ||
| # refer to https://github.com/aws/deep-learning-containers/blob/master/vllm/CHANGELOG.md#0110---2025-10-08 | ||
| repository: 763104351884.dkr.ecr.us-east-1.amazonaws.com/vllm |
Contributor
There was a problem hiding this comment.
For the blueprints, we should use the public ECR and not to force a region, we can give example how to optimize pulling by using DLC's private ECR repo, but in general, we should opt for the public ECR in the general examples.
https://gallery.ecr.aws/deep-learning-containers/vllm
Contributor
|
We now have the public ECR for DLC on all the vLLM images, please use that, also please reopen this PR in awslabs/ai-on-eks-charts as we now have a proper helm repository for this |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
What does this PR do?
For distributed inferencing using LWS, Amazon EFA provide a low latency inter-node connection that improves the latency of the inference. This PR provide an option to utilise Amazon EFA along with the Amaon DLC container images.
DLC images comes pre-baked with EFA libraries which are also added with the example
Motivation
LWS with Amzon EFA provide a way for multiple model layers to communicate over a low latency network resulting in better latency,
More
I have added the unit test for my changes in the repo too.
website/docsorwebsite/blogsection for this featurepre-commit run -awith this PR. Link for installing pre-commit locallyFor Moderators
Additional Notes