[https://nvbugs/6222480][fix] Fix stress

[xinhe-nv]

Summary by CodeRabbit

• Tests
  • Reduced stress test load parameters
  • Adjusted parameterized test case selection and coverage
  • Removed a waiver entry from the test configuration

Description

Test Coverage

PR Checklist

Please review the following before submitting your PR:

• PR description clearly explains what and why. If using CodeRabbit's summary, please make sure it makes sense.

• PR Follows TRT-LLM CODING GUIDELINES to the best of your knowledge.

• Test cases are provided for new code paths (see test instructions)

• If PR introduces API changes, an appropriate PR label is added - either api-compatible or api-breaking. For api-breaking, include BREAKING in the PR title.

• Any new dependencies have been scanned for license and vulnerabilities

• CODEOWNERS updated if ownership changes

• Documentation updated as needed

• Update tava architecture diagram if there is a significant design change in PR.

• The reviewers assigned automatically/manually are appropriate for the PR.

• Please check this after reviewing the above items as appropriate for this PR.

GitHub Bot Help

To see a list of available CI bot commands, please comment /bot help.

[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: d65be47f-1dac-498c-b749-04c6916f8cbb

📥 Commits

Reviewing files that changed from the base of the PR and between 0f5dc5a and e4e3b82.

📒 Files selected for processing (3)

• tests/integration/defs/disaggregated/test_disaggregated.py
• tests/integration/test_lists/qa/llm_function_rtx6k.txt
• tests/integration/test_lists/waives.txt

💤 Files with no reviewable changes (1)

• tests/integration/test_lists/waives.txt

---

📝 Walkthrough

Walkthrough

This PR reduces the disaggregated stress test request load from 60,000 to 30,000, removes the corresponding waiver indicating the lower load is safe, and rebalances the QA test selection list by disabling numerous DeepSeekV3Lite and GPTOSS test variants.

Changes

Test configuration and selection optimization

Layer / File(s)	Summary
Stress test request reduction and waiver removal tests/integration/defs/disaggregated/test_disaggregated.py, tests/integration/test_lists/waives.txt	The gpt_oss_120b_triton_stress test configuration reduces request_count from 60,000 to 30,000, and the associated H100 test waiver is removed, indicating the reduced load is approved for execution.
QA test selection rebalancing tests/integration/test_lists/qa/llm_function_rtx6k.txt	Multiple test variants are disabled in the QA list, including DeepSeekV3Lite test_bfloat16 with cuda_graph=True, bfloat16_4gpus_online_eplb, guided_decoding variants, and numerous nvfp4 parameter combinations, along with GPTOSS w4_* variants.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~10 minutes

Possibly related PRs

• NVIDIA/TensorRT-LLM#14721: Both PRs target the same H100 disaggregated test_disaggregated_stress_test stress variant (gpt_oss_120b_triton_stress), changing its test selection/waiver and/or parametrization to address the same stress-test issue.
• NVIDIA/TensorRT-LLM#14929: The main PR's update to tests/integration/test_lists/qa/llm_function_rtx6k.txt overlaps directly with PR #14929's changes to the same test-selection file (enabling/disabling the same DeepSeekV3Lite test_bfloat16 variants).

Suggested reviewers

• dongfengy
• StanleySun639

🚥 Pre-merge checks | ✅ 3 | ❌ 2

❌ Failed checks (1 warning, 1 inconclusive)

Check name	Status	Explanation	Resolution
Description check	⚠️ Warning	The PR description is entirely empty—only the template structure remains with no actual content filling any of the required sections.	Fill in the Description section explaining the stress issue being fixed, and populate the Test Coverage section with relevant test details.
Title check	❓ Inconclusive	The title is vague and lacks specificity about what stress issue is being fixed or which component/test is affected.	Provide a more descriptive title explaining what stress issue is fixed (e.g., 'Reduce stress test request count for gpt_oss_120b_triton' or 'Fix H100 disaggregated stress test load').

✅ Passed checks (3 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.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

⚔️ Resolve merge conflicts

• Resolve merge conflict in branch fix-stress

---

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

[xinhe-nv]

/bot run --stage-list ""

[tensorrt-cicd]

PR_Github #52031 [ run ] triggered by Bot. Commit: 9d307d0 Link to invocation

[tensorrt-cicd]

PR_Github #52031 [ run ] completed with state SUCCESS. Commit: 9d307d0
/LLM/main/L0_MergeRequest_PR pipeline #41369 (Partly Tested) completed with status: 'SUCCESS'

CI Report

Link to invocation

[xinhe-nv]

/bot reuse-pipeline

[tensorrt-cicd]

PR_Github #52040 [ reuse-pipeline ] triggered by Bot. Commit: 17efbd0 Link to invocation

[tensorrt-cicd]

PR_Github #52040 [ reuse-pipeline ] completed with state SUCCESS. Commit: 17efbd0
Reusing PR_Github #52031 (Partly Tested) for commit 17efbd0

Link to invocation