[https://nvbugs/6428144][test] Unwaive GB300 DeepSeek-R1 gen-only disagg PerfSanity

[chienchunhung]

Summary

Remove the waiver filed under NVBug 6428144 and restore post-merge CI coverage for:

perf/test_perf_sanity.py::test_e2e[disagg_upload-gen-only-gb300_deepseek-r1-fp4_128k8k_con256_ctx1_pp4_gen1_dep8_eplb0_mtp1_ccb-NIXL]

Rationale

PR #15737 normalized the gathered per-rank status before the gen-only disaggregated benchmark fill-gate reduction, addressing the list/int failure reported by NVBug 6438586. This waived workload exercises that gate and now passes on the fixed code.

NVBug 6428144 remains the tracker attached to this test's waiver; NVBug 6438586 and PR #15737 provide the related product-fix context.

Validation

Tested commit: 907c581354f5dc394a01fb5e8739270a281570be

• Targeted L0 pipeline #47307, launched by PR_Github #58724: the exact GB300 test was selected and passed (1 passed; 3,582.82-second call; pytest exit code 0).
• Full L0 pipeline #47430, launched by PR_Github #58888: passed on the same commit. The pipeline reused both successful targeted stages from #47307 and separately ran the CBTS-selected coverage (15 passed, 0 failed, 0 skipped).
• python3 scripts/check_test_list.py --validate --check-duplicate-waives — 2,209 unique entries validated.
• git diff --check

[chienchunhung]

/bot run --disable-fail-fast --stage-list "DGX_B200-16_GPUs-2_Nodes-PyTorch-Disagg-PerfSanity-CTX1-NODE1-GPU4-GEN1-NODE1-GPU8-Post-Merge-3,GB300-12_GPUs-3_Nodes-PyTorch-Disagg-PerfSanity-CTX1-NODE1-GPU4-GEN1-NODE2-GPU8-Post-Merge-3"

[coderabbitai]

📝 Walkthrough

Walkthrough

Removes one waiver entry from the integration test waiver list for the disagg_upload-gen_only-gb300_deepseek-r1-fp4 performance test configuration. No public entities or other waiver entries change.

Changes

Performance waiver update

Layer / File(s)	Summary
Remove targeted performance waiver tests/integration/test_lists/waives.txt	Deletes the perf/test_perf_sanity.py::test_e2e waiver for the specified GB300 DeepSeek configuration.

Estimated code review effort: 1 (Trivial) | ~2 minutes

Possibly related PRs

• NVIDIA/TensorRT-LLM#16093: Modifies related performance-test waiver entries in the same file.
• NVIDIA/TensorRT-LLM#16109: Touches similar disaggregated upload GB300 DeepSeek performance waivers.
• NVIDIA/TensorRT-LLM#16169: Changes waiver entries for the same performance test path.

Suggested reviewers: mzweilz, dpitman-nvda

🚥 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.
Title check	✅ Passed	The title is concise and accurately reflects removing the GB300 DeepSeek-R1 perf sanity waiver.
Description check	✅ Passed	The description explains the change, rationale, and validation clearly, with only the checklist section missing.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

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

[tensorrt-cicd]

PR_Github #58722 [ run ] triggered by Bot. Commit: 907c581 Link to invocation

[tensorrt-cicd]

PR_Github #58723 [ run ] triggered by Bot. Commit: 907c581 Link to invocation

[tensorrt-cicd]

PR_Github #58724 [ run ] triggered by Bot. Commit: 907c581 Link to invocation

[tensorrt-cicd]

PR_Github #58722 [ run ] completed with state ABORTED. Commit: 907c581

Link to invocation

[tensorrt-cicd]

PR_Github #58723 [ run ] completed with state ABORTED. Commit: 907c581

Link to invocation

[tensorrt-cicd]

PR_Github #58724 [ run ] completed with state SUCCESS. Commit: 907c581
/LLM/main/L0_MergeRequest_PR pipeline #47307 (Partly Tested) completed with status: 'SUCCESS'

CI Report

Link to invocation

[chienchunhung]

/bot run --disable-fail-fast

[tensorrt-cicd]

PR_Github #58888 [ run ] triggered by Bot. Commit: 907c581 Link to invocation

[tensorrt-cicd]

PR_Github #58888 [ run ] completed with state SUCCESS. Commit: 907c581
/LLM/main/L0_MergeRequest_PR pipeline #47430 completed with status: 'SUCCESS'

CI Report

Link to invocation

[chienchunhung]

/bot run --disable-fail-fast

[tensorrt-cicd]

PR_Github #59006 [ run ] triggered by Bot. Commit: 907c581 Link to invocation

[tensorrt-cicd]

PR_Github #59006 [ run ] completed with state SUCCESS. Commit: 907c581
/LLM/main/L0_MergeRequest_PR pipeline #47535 completed with status: 'SUCCESS'

CI Report

Link to invocation