1. Severity of the issue: (select one)
None: I’m just curious or want clarification.
Low: Annoying but doesn’t hinder my work.
Medium: Significantly affects my productivity but can find a workaround.
High: Completely blocks me.
2. Environment:
- Ray version: 2.20.0
- Python version: 3.11.9
- OS: Ubuntu 22.04.5 LTS
- Cloud/Infrastructure: Azure
- Other libs/tools (if relevant): conda
3. What happened vs. what you expected:
- Expected: Ray installs ray wheel from custom Python package index (Artifactory)
- Actual: Ray ignores pip configuration and attempts to install ray from ray-wheels.s3-us-west-2.amazonaws.com
For full transparency: I already posted this question on SO before I learned about the existence of this forum.
When submitting a Ray job using a conda runtime env (runtime_env = {"conda": "environment.yml"}
), Ray attempts to install the ray
wheel from ray-wheels.s3-us-west-2.amazonaws.com
even when I configure pip to use a different Python package index (--index-url
). Example environment.yml
:
name: test
channels:
- https://conda.anaconda.org/conda-forge/
- nodefaults
dependencies:
- pandas=2.2.3
- pip
- pip:
- --index-url https://artifactory.company.com/artifactory/api/pypi/pypi-remote/simple/
- ray==2.20.0
Ray seems to ignore the provided configuration and attempt to install the ray
wheel from ray-wheels.s3-us-west-2.amazonaws.com
. Is there a configuration setting to change this behavior or how can ray
be installed from a custom Python package index without having to go out to the Internet?