Skip to content

[spark-rapids] Fix cuda ppa conflict kernel614#1380

Open
nvliyuan wants to merge 1 commit intoGoogleCloudDataproc:mainfrom
nvliyuan:fix-cuda-ppa-conflict-kernel614
Open

[spark-rapids] Fix cuda ppa conflict kernel614#1380
nvliyuan wants to merge 1 commit intoGoogleCloudDataproc:mainfrom
nvliyuan:fix-cuda-ppa-conflict-kernel614

Conversation

@nvliyuan
Copy link
Contributor

@nvliyuan nvliyuan commented Mar 11, 2026

this pr is to :
1:fix issue on dp3-ubuntu image
2:update hot fix release spark-rapids version v26.02.1

@gemini-code-assist
Copy link
Contributor

Summary of Changes

Hello, 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 addresses and resolves an issue related to CUDA PPA conflicts, specifically observed on dp3-ubuntu images with newer kernel versions. It updates the default Spark RAPIDS version and significantly refines the GPU driver installation script to exclusively leverage the NVIDIA official repository for CUDA and driver components, thereby eliminating dependencies on the graphics-drivers PPA and ensuring smoother installations.

Highlights

  • Spark RAPIDS Version Update: The default Spark RAPIDS version has been updated from 25.12.0 to 26.02.1.
  • GPU Driver Installation Refinement: The GPU driver installation process for repository-based installs has been modified to remove the use of the graphics-drivers PPA, resolving potential conflicts with newer kernels.
  • Streamlined CUDA Installation: CUDA toolkit and NVIDIA drivers are now installed using the unified cuda meta-package directly from the NVIDIA official repository, simplifying the process and avoiding PPA-related issues.

🧠 New Feature in Public Preview: You can now enable Memory to help Gemini Code Assist learn from your team's feedback. This makes future code reviews more consistent and personalized to your project's style. Click here to enable Memory in your admin console.

Changelog
  • spark-rapids/spark-rapids.sh
    • Updated the DEFAULT_SPARK_RAPIDS_VERSION from "25.12.0" to "26.02.1".
    • Removed the steps for adding the ppa:graphics-drivers/ppa repository during repository-based GPU driver installation.
    • Modified the CUDA and NVIDIA driver installation commands to install the cuda meta-package from the NVIDIA official repository, replacing separate cuda-toolkit and nvidia-driver installations.
    • Added comments to clarify the change to use only the NVIDIA official repository to prevent PPA conflicts.
Activity
  • No specific activity (comments, reviews, progress updates) has been recorded for this pull request yet.
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 by creating a comment using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands on the current page.

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 pull request 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 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. If you're interested in giving your feedback about your experience with Gemini Code Assist for GitHub and other Google products, sign up here.

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.

@nvliyuan
Copy link
Contributor Author

CC @viadea

Copy link
Contributor

@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 updates the Spark RAPIDS installation script to resolve a PPA conflict on newer Ubuntu kernels by removing the graphics-drivers PPA and using NVIDIA's cuda metapackage, which simplifies the installation. The default Spark RAPIDS version is also updated. My review has found a critical issue with the new version number, which appears to be a typo and would break the script.


# Update SPARK RAPIDS config
readonly DEFAULT_SPARK_RAPIDS_VERSION="25.12.0"
readonly DEFAULT_SPARK_RAPIDS_VERSION="26.02.1"
Copy link
Contributor

Choose a reason for hiding this comment

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

critical

The RAPIDS version 26.02.1 appears to be a typo. The RAPIDS versioning scheme is YY.MM.patch, so this version refers to a future date (February 2026) and will likely cause the script to fail when trying to download the JAR. This was probably intended to be 24.02.1. Note that the previous version 25.12.0 was also likely a typo for a past version (e.g., 23.12.0).

Suggested change
readonly DEFAULT_SPARK_RAPIDS_VERSION="26.02.1"
readonly DEFAULT_SPARK_RAPIDS_VERSION="24.02.1"

@nvliyuan
Copy link
Contributor Author

image

- Remove graphics-drivers PPA to avoid package conflicts with NVIDIA official repo
- Use 'cuda' meta-package instead of separate cuda-toolkit and nvidia-driver installation
- The cuda package automatically includes compatible toolkit and driver versions

This fixes the "held broken packages" error when installing nvidia-driver-570-open
due to conflicts between PPA packages and NVIDIA official repository packages.

Signed-off-by: liyuan <yuali@nvidia.com>
Made-with: Cursor
@nvliyuan nvliyuan force-pushed the fix-cuda-ppa-conflict-kernel614 branch from 344a4d7 to 4ce0529 Compare March 11, 2026 03:14
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