inference icon indicating copy to clipboard operation
inference copied to clipboard

publish roboflow/roboflow-inference-server-cpu to AWS ECR in addition to Dockerhub

Open elijahchancey-arena opened this issue 9 months ago • 2 comments

Search before asking

  • [x] I have searched the Inference issues and found no similar feature requests.

Description

I'm trying to deploy roboflow/roboflow-inference-server-cpu via ECS Fargate on AWS. Because I am not a Dockerhub customer, I am experiencing rate limits from a low number of image pulls per hour.

If you publish roboflow/roboflow-inference-server-cpu to the public AWS ECR gallery, I will not be subject to the rate limits.

As a workaround, I am manually publishing roboflow/roboflow-inference-server-cpu to a private ECR repository.

Use case

This will allow teams to pull your docker images from AWS ECR and not be subject to Dockerhub rate limits.

Additional

No response

Are you willing to submit a PR?

  • [ ] Yes I'd like to help by submitting a PR!

elijahchancey-arena avatar Mar 20 '25 20:03 elijahchancey-arena

hi @elijahchancey-arena

We are checking with docker hub because I think we have it setup so you shouldn't hit rate limits for our these docker images (via our paid docker account), but might have something misconfigured or I have understood something wrong. If we can't disable the rate limits on docker hub publicly, we'll certainly consider this.

hansent avatar Mar 31 '25 18:03 hansent

Thank you.

elijahchancey-arena avatar Apr 01 '25 14:04 elijahchancey-arena