fix(webhook): count the resource to fit quota in webhook when reqnum > 1

[DSFans2014]

What type of PR is this?

/kind bug

What this PR does / why we need it:

count the resource to fit quota in webhook if reqnum > 1

Which issue(s) this PR fixes:
Fixes #

Special notes for your reviewer:

Does this PR introduce a user-facing change?:

Summary by CodeRabbit

• Bug Fixes
  • Improved GPU quota checks so memory and CPU requests are correctly counted when a workload requests multiple GPUs.
  • Multi-GPU jobs with matching memory requests are now more likely to pass resource fit validation as expected.
  • Added coverage for a multi-GPU scheduling scenario to help prevent regressions.

[coderabbitai]

📝 Walkthrough

Walkthrough

The quota-fit logic in fitResourceQuota was changed to scale accumulated memory and core requests by the requested GPU count instead of only accumulating when count equals 1. A new test case validates multi-GPU requests fitting within quota.

Changes

Multi-GPU Quota Fix

Layer / File(s)	Summary
Scaled quota accumulation and test coverage pkg/scheduler/webhook.go, pkg/scheduler/webhook_test.go	fitResourceQuota now multiplies memory/core requests by the count instead of restricting accumulation to count==1; a new test case verifies a pod requesting 2 GPUs with 400 GPUMem fits within quota.

Estimated code review effort: 2 (Simple) | ~10 minutes

Poem

A hop, a scale, a count times two,
Quotas now compute what's true.
No more stuck at "req == 1",
Multi-GPU math is finally done!
🐇✨ Thump-thump, tests all run green.

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
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.
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title matches the main bug fix: webhook quota fitting now counts requested resources when reqnum > 1.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

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

[gemini-code-assist]

Code Review

This pull request updates the resource quota fitting logic in the scheduler webhook to correctly scale memory and core requirements when multiple resources (such as GPUs) are requested, rather than assuming a single resource request. A corresponding test case has been added to verify this behavior. There are no review comments, so I have no feedback to provide.

Important

The consumer version of Gemini Code Assist on GitHub is being sunset. Starting June 18, 2026, new organization installations will be blocked, and all code review activity will officially cease on July 17, 2026.
For more details on the timeline and next steps, please review the Help Documentation.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

Flag	Coverage Δ
unittests	59.09% <100.00%> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

Files with missing lines	Coverage Δ
pkg/scheduler/webhook.go	76.19% <100.00%> (ø)

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[mesutoezdil]

/lgtm

[mesutoezdil]

the old code assumed reqnum was always 1 — so if a container requested 2 GPUs with 400MB memory each, it would only count 400MB against quota instead of 800MB. multiplying by req fixes that.

same issue existed for cores. good catch.

Why are you rewriting the existing PR definition and code as a comment? Is this some kind of motivational support?

[DSFans2014]

the old code assumed reqnum was always 1 — so if a container requested 2 GPUs with 400MB memory each, it would only count 400MB against quota instead of 800MB. multiplying by req fixes that.
same issue existed for cores. good catch.

Why are you rewriting the existing PR definition and code as a comment? Is this some kind of motivational support?

@mesutoezdil Where does your quoted content come from?

[archlitchi]

/lgtm

[hami-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: archlitchi, DSFans2014

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Details

Needs approval from an approver in each of these files:

• OWNERS [DSFans2014,archlitchi]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[mesutoezdil]

the old code assumed reqnum was always 1 — so if a container requested 2 GPUs with 400MB memory each, it would only count 400MB against quota instead of 800MB. multiplying by req fixes that.
same issue existed for cores. good catch.

Why are you rewriting the existing PR definition and code as a comment? Is this some kind of motivational support?

@mesutoezdil Where does your quoted content come from?

As usual, someone had pasted a nonsensical LLM output; after i’d replied, he deleted the post.

We are working on it: #1998 (comment)

Answers should be written by human.