ci: build container on Azure for main builds#2521
Draft
kajalj22 wants to merge 1 commit into
Draft
Conversation
|
Auto-sync is disabled for ready for review pull requests in this repository. Workflows must be run manually. Contributors can view more details about this message here. |
|
Auto-sync is disabled for draft pull requests in this repository. Workflows must be run manually. Contributors can view more details about this message here. |
Contributor
Author
|
/ok to test 9e8e669 |
kajalj22
force-pushed
the
kajalj/seed-azure-cache
branch
from
9e8e669 to
a74912a
Compare
Contributor
Author
|
/ok to test a74912a |
chtruong814
reviewed
May 18, 2026
|
|
||
| - name: Copy main image to Azure registry | ||
| env: | ||
| SRC: ${{ needs.org-member-pre-flight.outputs.registry }}/${{ vars.CI_CONTAINER_NAME }}:main |
Contributor
There was a problem hiding this comment.
What do you expect this to resolve to for external contributor?
Add a parallel build-container-azure job that builds and pushes the container image to the Azure registry (nemoci.azurecr.io) on every main branch build. This runs alongside the existing GCP build since the architectures differ (GCP=ARM64, Azure=x86_64) and images cannot be copied across. External contributor PR builds use Azure runners with inline cache from nemoci.azurecr.io/rl:main. Without this job, the Azure :main tag goes stale whenever the Dockerfile changes, causing cache misses that can fail with nvcr.io rate-limit 401 errors. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com> Signed-off-by: Kajal Jain <kajalj@nvidia.com>
kajalj22
force-pushed
the
kajalj/seed-azure-cache
branch
from
cc1099e to
2e6810a
Compare
kajalj22
changed the title
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
2 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.
Summary
build-container-azurejob tocicd-main.ymlthat builds and pushes the container image tonemoci.azurecr.ioon every main branch build (including nightly):maintag fresh so external contributor PR builds have a warm inline cache and can runCI:Lfast/CI:docs(which reuse the:mainimage)Context
External contributor builds run on Azure x86_64 runners and use
nemoci.azurecr.io/rl:mainas an inline Docker cache source. Since main branch builds only ran on GCP ARM64 runners, the Azure:maintag was never updated — causing two issues:cuda-dl-base:26.03-cuda13.2in chore: Enable cuda-13 build #2332, BuildKit couldn't find cached layers and tried pulling from nvcr.io directly, hitting rate-limit 401 errors (e.g. feat: add AIME-2026 benchmark. #2469)CI:Lfast/CI:docsbroken for external contributors: These test levels skip the container build and reuse the:mainimage, which didn't exist on the Azure registryTest plan
build-container-azureruns on next nightly/main pushCI:Lfastworks for an external contributor PR🤖 Generated with Claude Code