refactor(citestwheel): switch base image from devel to base - #408
Merged
Conversation
rapids-bot Bot
pushed a commit
that referenced
this pull request
May 19, 2026
Debugging failures that showed up in https://github.com/rapidsai/ci-imgs/actions/runs/26057524971/job/76609236506?pr=408#step:9:6848 while working on #408 @jameslamb is seeing the same thing over in `rapidsai/docker` (tracked at rapidsai/docker#878) **update 2026-05-19** @jameslamb figured out that force-reinstalling `conda` before performing any other `conda` operations fixes the notices loading issue. I've applied those same changed here and reverted the previous change. Authors: - Gil Forsyth (https://github.com/gforsyth) Approvers: - Kyle Edwards (https://github.com/KyleFromNVIDIA) - James Lamb (https://github.com/jameslamb) URL: #409
gforsyth
force-pushed
the
switch_devel_to_base
branch
from
May 19, 2026 19:22
c69238c to
d0fa8d9
Compare
gforsyth
force-pushed
the
switch_devel_to_base
branch
from
June 30, 2026 13:47
d0fa8d9 to
a3edbfe
Compare
Contributor
|
After we merge this, let's work on rapidsai/docs#673 and use |
jameslamb
approved these changes
Jul 21, 2026
Member
There was a problem hiding this comment.
awesome!!!
The images look like they're about 10x smaller (400ish MB instead of 4ish GB) and I saw container spin-up times around 30 seconds on the testing PRs (usually like 3-4 minutes today).
Let's merge this. The testing looks pretty thorough, I'm around to help with whatever other things may break because of this.
7 tasks
Contributor
|
That's an awesome speedup. Glad we're finally in a position to do this! |
Contributor
Author
|
Let's goooooooooooooo |
Contributor
Author
|
/merge |
3 tasks
rapids-bot Bot
pushed a commit
to NVIDIA/cudf
that referenced
this pull request
Jul 22, 2026
The `ci-imgs` change in rapidsai/ci-imgs#408 switched `citestwheel` images from CUDA `devel` to `base`, which no longer bundles `libcufile.so` with the container. Now we must install cuFile wheels in Python test environments that use CUDA wheels. CUDA 12 uses `nvidia-cufile-cu12` with some caveats in the pinnings, because cuFile wheels weren't available until 12.6.3 so the `[cufile]` extra isn't reliable. CUDA 13 uses the `cuda-toolkit[cufile]` extra. Authors: - Bradley Dice (https://github.com/bdice) - Gil Forsyth (https://github.com/gforsyth) Approvers: - Gil Forsyth (https://github.com/gforsyth) URL: #23392
3 tasks
rapids-bot Bot
pushed a commit
to NVIDIA/cudf
that referenced
this pull request
Jul 30, 2026
It looks like up until this point our usage of numba-cuda from CUDA wheels has been relying on the system CCCL rather than the one bundled with libcudf. We noticed this as a result of the change in ci-imgs to use the base rather than devel CUDA images as our base (rapidsai/ci-imgs#408) because testing on older CTKs revealed that headers required by numba-cuda were missing (specifically `cuda/atomic`). To ensure that we have ABI-compatible usage of CCCL given that we have custom numba extension code for our string UDFs, we're best off ensuring that we use the CCCL bundled in libcudf rather than whatever is found on the system, so this PR adds that directory to our numba search path. Authors: - Vyas Ramasubramani (https://github.com/vyasr) Approvers: - James Lamb (https://github.com/jameslamb) - Matthew Roeschke (https://github.com/mroeschke) URL: #23455
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
4 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.
New PR because I rebased and force-pushed -- rebuilding
ci-imgsagainst cudabaseDepends on rapidsai/build-planning#279.
Testing is tracked in rapidsai/build-planning#143 (comment)
As of 2026-07-21 -- all libraries that load shared objects from wheel installs are tested against the cuda-base image, and all fixes to ensure that they correctly load those shared objects have landed in their respective projects.