Add CUDA image variants and smoke tests

[aapdo]

Summary

• Add build variants for CUDA 11.8, 12.2, 12.5, and 12.8 with matching TensorFlow versions.
• Add Miniforge/micromamba, JupyterLab, system packages, and noVNC/VNC runtime support to the image flow.
• Add image variant build helpers and Ansible-based smoke tests for GPU visibility, TensorFlow GPU detection, Jupyter, noVNC, and uid/create_container.sh compatibility.

Validation

• Ran local syntax and config checks for Docker/entrypoint/test helper changes.
• Ran remote smoke tests on LAB8, LAB9, LAB10, FARM1, and FARM9.
• Built and pushed Docker Hub images under dguailab/decs for CUDA 11.8, 12.2, 12.5, and 12.8 variants.

[gitguardian]

⚠️ GitGuardian has uncovered 1 secret following the scan of your pull request.

Please consider investigating the findings and remediating the incidents. Failure to do so may lead to compromising the associated services or software components.

🔎 Detected hardcoded secret in your pull request

GitGuardian id	GitGuardian status	Secret	Commit	Filename
32875495	Triggered	Generic Password	a45066a	scripts/test_uid_create_container.py	View secret

🛠 Guidelines to remediate hardcoded secrets

1. Understand the implications of revoking this secret by investigating where it is used in your code.
2. Replace and store your secret safely. Learn here the best practices.
3. Revoke and rotate this secret.
4. If possible, rewrite git history. Rewriting git history is not a trivial act. You might completely break other contributing developers' workflow and you risk accidentally deleting legitimate data.

To avoid such incidents in the future consider

• following these best practices for managing and storing secrets including API keys and other credentials
• install secret detection on pre-commit to catch secret before it leaves your machine and ease remediation.

---

^{🦉 GitGuardian detects secrets in your source code to help developers and security teams secure the modern development process. You are seeing this because you or someone else with access to this repository has authorized GitGuardian to scan your pull request.}

[Copilot]

Pull request overview

This PR re-architects the DECS Docker image build to support multiple CUDA/TensorFlow variants, replacing the single-image flow built on tensorflow/tensorflow:2.18.0-gpu with a parameterized Dockerfile driven by image-variants.json. It adds Miniforge/JupyterLab/noVNC runtime support, refactors the entrypoint with driver/CUDA compatibility checks and an opt-in VNC stack, updates the GitHub Actions workflow to fan out builds via a matrix, and introduces Python helpers plus Ansible playbooks for building and smoke-testing the variants on remote GPU hosts.

Changes:

• Parameterized Dockerfile + image-variants.json + matrix-based GitHub Actions workflow for CUDA 11.8/12.2/12.5/12.8 builds.
• New entrypoint.sh with image runtime info, optional STRICT_CUDA_COMPAT driver check, and TigerVNC/noVNC startup.
• New scripts/ (build/test/uid helpers) and tests/ansible/ playbooks (build + smoke), with documentation rewritten in README.md.

Reviewed changes

Copilot reviewed 12 out of 13 changed files in this pull request and generated 6 comments.

Show a summary per file

File	Description
.dockerignore	Excludes git, scripts, tests, and README from the build context.
.github/workflows/docker-publish.yml	Adds a prepare job that emits a build matrix and updates build-and-push to consume per-variant build args.
.gitignore	Ignores Python bytecode artifacts.
Dockerfile	Switches to nvidia/cuda base, parameterizes CUDA/TF/Python via build args, installs Miniforge/JupyterLab/VNC stack.
README.md	Documents the variant matrix, build/test commands, VNC env vars, and admin notes.
entrypoint.sh	Adds image runtime banner, driver compatibility gate, VNC/noVNC startup, and uses $JUPYTER_BIN/$CONDA_DIR.
image-variants.json	Declares the four supported CUDA/TF variants with aliases and minimum driver versions.
scripts/build_variants.py	CLI to build (and optionally push) variants from the manifest.
scripts/test_image_variants.py	Local CPU/GPU smoke runner per variant.
scripts/test_uid_create_container.py	Drives ~/uid/script_test/create_container.sh for each variant.
scripts/variant_matrix.py	Loads the manifest, validates uniqueness, and emits a GitHub Actions matrix.
tests/ansible/decs_image_build.yml	Uploads build context tarball and runs docker build on remote hosts.
tests/ansible/decs_image_smoke.yml	Runs the image with create_container.sh-style env, validates GPU/TF/Jupyter/noVNC.

Comments suppressed due to low confidence (1)

entrypoint.sh:12

• is_truthy accepts true/TRUE/1/yes/YES/on/ON but not the common Python-style True/False. Since environment variables in this repo flow through Ansible and Python helpers (e.g. the smoke playbook normalizes via ternary('true', 'false'), but ad-hoc docker run -e ENABLE_VNC=True from operators is plausible), consider also accepting mixed-case True/Yes/On (e.g., normalize with ${1,,} before matching). Otherwise ENABLE_VNC=True will be silently treated as false.

is_truthy() {
    case "${1:-}" in
        true|TRUE|1|yes|YES|on|ON) return 0 ;;
        *) return 1 ;;
    esac
}

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.