fix(nemo-deployments): release docker gpu for missing one-shot containers

[ironcommit]

Summary by CodeRabbit

• Bug Fixes
  • Fixed GPU resource handling when one-shot deployments lose their container.
  • One-shot deployments now correctly report a failed status and release allocated GPU resources.
  • Preserved existing behavior for deployments configured to remain lost when containers are missing.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Path: .coderabbit.yaml

Review profile: CHILL

Plan: Enterprise

Run ID: 4881619b-915c-4475-bb47-c7ae0759f02d

📥 Commits

Reviewing files that changed from the base of the PR and between e13178a and cd0816e.

📒 Files selected for processing (2)

• plugins/nemo-deployments/src/nemo_deployments_plugin/backends/docker/backend.py
• plugins/nemo-deployments/tests/unit/backends/docker/test_backend_mocked.py

---

📝 Walkthrough

Walkthrough

DockerDeploymentBackend.read_status now releases GPU allocations when one-shot deployments lose their containers, reuses the deployment key across exit paths, and adds tests for release and non-release cases.

Changes

Docker deployment status cleanup

Layer / File(s)	Summary
Status handling and regression coverage plugins/nemo-deployments/src/nemo_deployments_plugin/backends/docker/backend.py, plugins/nemo-deployments/tests/unit/backends/docker/test_backend_mocked.py	read_status computes dep_key before lookup, releases GPUs for missing Never and OnFailure containers, reuses the key for exit handling, and verifies terminal status and release behavior in mocked tests.

Possibly related PRs

• NVIDIA-NeMo/nemo-platform#868: Also changes missing-container status and GPU release handling in DockerDeploymentBackend.read_status.
• NVIDIA-NeMo/nemo-platform#925: Also restructures one-shot GPU release logic across missing-container and exit paths.

Suggested labels: test

Suggested reviewers: tylersbray, benmccown

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly matches the main change: releasing Docker GPU resources when one-shot containers are missing.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch docker-missing-container-gpu-release/rsadler

---

_{Comment @coderabbitai help to get the list of available commands.}

[github-actions]

Suite	Lines Covered	Line Rate	Branch Rate
Unit Tests	27565/35320	78.0%	62.5%
Integration Tests	16091/34032	47.3%	19.8%

[tylersbray]

Nice.