[Slurm] Set reasonable default CPU and memory for GPU instances

[kevinmingtarja]

For Kubernetes, we default to a reasonable size of CPU and memory when GPU is requested. However for Slurm, the CPU and memory is only 2 and 2 GB (same as CPU only instances).

This PR fixes it to have the same behaviour as Kubernetes.

Tested (run the relevant ones):

• Code formatting: install pre-commit (auto-check on commit) or bash format.sh
• Any manual or new tests for this PR (please specify below)
• All smoke tests: /smoke-test (CI) or pytest tests/test_smoke.py (local)
• Relevant individual tests: /smoke-test -k test_name (CI) or pytest tests/test_smoke.py::test_name (local)
• Backward compatibility: /quicktest-core (CI) or pytest tests/smoke_tests/test_backward_compat.py (local)

[gemini-code-assist]

Summary of Changes

Hello @kevinmingtarja, 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 resolves an inconsistency in how GPU instances are provisioned on Slurm clusters. Previously, GPU instances would receive the same minimal CPU and memory defaults as CPU-only instances. This change updates the Slurm cloud provider to automatically assign more appropriate and generous default CPU and memory resources (4 CPUs and 16GB memory per GPU) when a GPU is requested without explicit resource specifications, mirroring the established behavior in Kubernetes. This ensures that GPU workloads on Slurm are allocated sufficient resources by default, improving usability and performance.

Highlights

• Default GPU Resource Allocation for Slurm: Introduced reasonable default CPU and memory values for GPU instances in Slurm, aligning with Kubernetes behavior. Specifically, _DEFAULT_NUM_VCPUS_WITH_GPU is set to 4 and _DEFAULT_MEMORY_CPU_RATIO_WITH_GPU to 4.
• Slurm Cloud Provider Logic Update: Modified the _make method within sky/clouds/slurm.py to utilize these new default CPU and memory values when GPU resources are requested without explicit CPU or memory specifications.
• New Unit Tests for GPU Defaults: Added a new test class TestSlurmGPUDefaults in tests/unit_tests/test_sky/clouds/test_slurm.py to thoroughly verify the correct application of default CPU and memory for GPU instances, including scenarios with explicit overrides and different GPU counts.

🧠 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.

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. ↩

[gemini-code-assist]

Code Review

This pull request aligns Slurm's default CPU and memory allocation for GPU instances with Kubernetes' behavior, which is a sensible improvement. The changes in sky/clouds/slurm.py correctly implement the new logic using shared constants. The new tests in tests/unit_tests/test_sky/clouds/test_slurm.py are comprehensive, covering default behavior, overrides, and regressions. The implementation is clean, correct, and well-tested.

[kevinmingtarja]

/smoke-test --slurm -k minimal

[SeungjinYang]

Kinda wondering if it's worth factoring this logic out of both k8s and slurm, but we can also defer it to the next time that becomes a nuisance / problem