[https://nvbugs/6681216][infra] Waive DeepSeek V3 Lite NIXL on B200 and B300

[chienchunhung]

Description

Temporarily waive the following test on DGX B200 and DGX B300 while keeping its H100 coverage active:

disaggregated/test_disaggregated.py::test_disaggregated_deepseek_v3_lite_fp8_nixl[DeepSeek-V3-Lite-fp8]

On both Blackwell platforms, the workers remain at 0/2 registered while the logs reach Blackwell JIT and Running autotuner warmup.... The harness then raises TimeoutError: Timeout after 300s waiting for _check before inference or KV transfer begins.

The same failure was observed on main in L0_PostMerge #2933. In the CI history inspected after PR #17792 enabled this existing test on Blackwell, all 23 executed B200/B300 records failed and none passed; 12 additional records were skipped on stale branches. PR #17792 is the test-exposure boundary, but is not proven to have introduced the underlying startup problem.

This startup failure is tracked by NVBUG 6681216. It is separate from the KV-transfer timeout addressed by PR #17137 / NVBUG 6480621 and from the earlier host OOM addressed by PR #18185.

Validation

• python scripts/check_test_list.py --check-duplicate-waives --validate
• git diff --check

Both checks pass. The test-list validator checked 1,944 unique entries and found no duplicate waivers.

Dev Engineer Review

• Added temporary waiver entries for test_disaggregated_deepseek_v3_lite_fp8_nixl[DeepSeek-V3-Lite-fp8] on DGX B200 and DGX B300.
• Linked both entries to NVBUG 6681216.
• Retained H100 coverage.
• Waiver format and scope are consistent. No duplicate entries were found.
• git diff --check passed.

QA Engineer Review

• Modified tests/integration/test_lists/waives.txt.
• Added DGX B200 and DGX B300 waiver entries for the DeepSeek V3 Lite FP8 NIXL test.
• No test-db/ or qa/ files were modified.
• CBTS coverage data is unavailable.

Verdict: needs follow-up

[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: 11ea3f14-c77c-4118-a5aa-9a6bb6823ff7

📥 Commits

Reviewing files that changed from the base of the PR and between 732cd83 and cf15d9f.

📒 Files selected for processing (1)

• tests/integration/test_lists/waives.txt

Included review availability: Your plan provides up to 12 included reviews per hour; 11 remain after this review.

---

Walkthrough

The integration test waiver list now skips the DeepSeek V3 Lite FP8 NIXL disaggregated-serving test on DGX B200 and DGX B300. Both entries reference NVBug 6681216.

Changes

Test waivers

Layer / File(s)	Summary
Add DGX test waivers tests/integration/test_lists/waives.txt	Added skip entries for the DeepSeek V3 Lite FP8 NIXL disaggregated-serving test on DGX B200 and DGX B300. Both entries reference NVBug 6681216.

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

Merge Risk: ⚪ Minimal · up to cf15d

This change narrowly waives the affected Blackwell test while preserving H100 coverage, and the listed validation checks pass. No actionable merge-blocking risk remains after normal review.

Possibly related PRs

• NVIDIA/TensorRT-LLM#15195: Adds temporary test waiver entries to the same waiver list.

Suggested reviewers: brnguyen2

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly identifies the NVBug, uses the valid infra type, and states the main change: waiving the DeepSeek V3 Lite NIXL test on B200 and B300.
Description check	✅ Passed	The description clearly explains the failure, affected platforms, scope of the waiver, tracking bug, retained H100 coverage, and validation results. It does not use the template's exact Test Coverage …
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check. Docstring coverage is scoped to functions touched by this diff. Analyzed 0 functions across 0…
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.

Full details: Description check

Explanation

The description clearly explains the failure, affected platforms, scope of the waiver, tracking bug, retained H100 coverage, and validation results. It does not use the template's exact Test Coverage or PR Checklist headings, but it provides the relevant test and validation information and is otherwise complete.

Full details: Docstring Coverage

Explanation

No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check. Docstring coverage is scoped to functions touched by this diff. Analyzed 0 functions across 0 files. (1 skipped: 1 unsupported.)

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

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

[chienchunhung]

/bot skip --comment "Test waiving only"

[brb-nv]

LGTM.

[tensorrt-cicd]

PR_Github #69815 [ skip ] triggered by Bot. Commit: cf15d9f Link to invocation

[tensorrt-cicd]

PR_Github #69815 [ skip ] completed with state SUCCESS. Commit: cf15d9f
Skipping testing for commit cf15d9f

Link to invocation