Skip to content

[https://nvbugs/6302903][infra] Unwaive deepseek-r1-fp4 disagg perf-sanity e2e test - #15787

Closed
xwang233 wants to merge 3 commits into
NVIDIA:mainfrom
xwang233:unwaive-6302903-deepseek-r1-disagg-perfsanity
Closed

[https://nvbugs/6302903][infra] Unwaive deepseek-r1-fp4 disagg perf-sanity e2e test#15787
xwang233 wants to merge 3 commits into
NVIDIA:mainfrom
xwang233:unwaive-6302903-deepseek-r1-disagg-perfsanity

Conversation

@xwang233

@xwang233 xwang233 commented Jun 30, 2026

Copy link
Copy Markdown
Collaborator

Unwaive perf/test_perf_sanity.py::test_e2e[disagg_upload-e2e-gb200_deepseek-r1-fp4_8k1k_con4096_ctx1_dep4_gen1_dep16_eplb0_mtp1_ccb-NIXL] (nvbugs/6302903).

The test passes in a local reproduction run on GB200 (0/20480 failed requests), so the waiver is removed to let CI exercise it again.

Summary by CodeRabbit

  • Chores
    • Removed an obsolete test waiver from the integration test list.

…anity e2e test

Signed-off-by: Xiao Wang <24860335+xwang233@users.noreply.github.com>
@coderabbitai

coderabbitai Bot commented Jun 30, 2026

Copy link
Copy Markdown
Contributor

Review Change Stack

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: cbe586bd-d472-4bbf-aba5-fb53c4339bb5

📥 Commits

Reviewing files that changed from the base of the PR and between 23c835d and afada7d.

📒 Files selected for processing (1)
  • tests/integration/test_lists/waives.txt
💤 Files with no reviewable changes (1)
  • tests/integration/test_lists/waives.txt

📝 Walkthrough

Walkthrough

Removes one SKIP waiver entry from tests/integration/test_lists/waives.txt for a specific perf/test_perf_sanity.py::test_e2e disagg_upload configuration on GB200 with DeepSeek-R1-FP4, previously waived via nvbugs/6302903.

Changes

Waiver Removal

Layer / File(s) Summary
Remove perf test waiver entry
tests/integration/test_lists/waives.txt
Deletes the SKIP waiver for perf/test_perf_sanity.py::test_e2e[disagg_upload-e2e-gb200_deepseek-r1-fp4_8k1k_con4096_ctx1_dep4_gen1_dep16_eplb0_mtp1_ccb-NIXL] (nvbugs/6302903).

Estimated code review effort

🎯 1 (Trivial) | ⏱️ ~2 minutes

Possibly related PRs

Suggested reviewers

  • jieli-matrix
  • xinhe-nv
  • yingguo-trt
🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name Status Explanation Resolution
Description check ⚠️ Warning The description explains why the waiver was removed, but it omits the required template sections for Description, Test Coverage, and PR Checklist. Add the required template sections, especially a brief Description, Test Coverage, and completed PR Checklist items.
✅ Passed checks (4 passed)
Check name Status Explanation
Title check ✅ Passed The title matches the change: it identifies the nvbugs ticket and clearly says the perf-sanity test is being unwaived.
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.
✨ Finishing Touches
🧪 Generate unit tests (beta)
  • Create PR with unit tests

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

@xwang233

Copy link
Copy Markdown
Collaborator Author

/bot run --add-multi-gpu-test --stage-list "GB200-20_GPUs-5_Nodes-PyTorch-Disagg-PerfSanity-CTX1-NODE1-GPU4-GEN1-NODE4-GPU16-Post-Merge"

@tensorrt-cicd

Copy link
Copy Markdown
Collaborator

PR_Github #56668 [ run ] triggered by Bot. Commit: afada7d Link to invocation

@tensorrt-cicd

Copy link
Copy Markdown
Collaborator

PR_Github #56668 [ run ] completed with state FAILURE. Commit: afada7d
/LLM/main/L0_MergeRequest_PR pipeline #45492 (Partly Tested) completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

  • Please check the failed tests and fix your PR
  • If you cannot view the failures, ask the CI triggerer to share details
  • Once fixed, request an NVIDIA team member to trigger CI again

CI Agent Failure Analysis

Link to invocation

@xwang233

Copy link
Copy Markdown
Collaborator Author

/bot run --add-multi-gpu-test --stage-list "GB200-20_GPUs-5_Nodes-PyTorch-Disagg-PerfSanity-CTX1-NODE1-GPU4-GEN1-NODE4-GPU16-Post-Merge"

@tensorrt-cicd

Copy link
Copy Markdown
Collaborator

PR_Github #56686 [ run ] triggered by Bot. Commit: afada7d Link to invocation

@tensorrt-cicd

Copy link
Copy Markdown
Collaborator

PR_Github #56686 [ run ] completed with state FAILURE. Commit: afada7d
/LLM/main/L0_MergeRequest_PR pipeline #45509 (Partly Tested) completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

  • Please check the failed tests and fix your PR
  • If you cannot view the failures, ask the CI triggerer to share details
  • Once fixed, request an NVIDIA team member to trigger CI again

CI Agent Failure Analysis

Link to invocation

…eek-r1-disagg-perfsanity

Signed-off-by: Xiao Wang <xiaow@nvidia.com>
Signed-off-by: Xiao Wang <24860335+xwang233@users.noreply.github.com>
@xwang233

Copy link
Copy Markdown
Collaborator Author

/bot run --stage-list "GB200-20_GPUs-5_Nodes-PyTorch-Disagg-PerfSanity-CTX1-NODE1-GPU4-GEN1-NODE4-GPU16-Post-Merge-*"

@tensorrt-cicd

Copy link
Copy Markdown
Collaborator

PR_Github #56752 [ run ] triggered by Bot. Commit: ceb1189 Link to invocation

@tensorrt-cicd

Copy link
Copy Markdown
Collaborator

PR_Github #56752 [ run ] completed with state FAILURE. Commit: ceb1189
/LLM/main/L0_MergeRequest_PR pipeline #45572 (Partly Tested) completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

  • Please check the failed tests and fix your PR
  • If you cannot view the failures, ask the CI triggerer to share details
  • Once fixed, request an NVIDIA team member to trigger CI again

CI Agent Failure Analysis

Link to invocation

@xwang233

xwang233 commented Jul 1, 2026

Copy link
Copy Markdown
Collaborator Author

/bot run --stage-list "GB200-20_GPUs-5_Nodes-PyTorch-Disagg-PerfSanity-CTX1-NODE1-GPU4-GEN1-NODE4-GPU16-Post-Merge-*"

@tensorrt-cicd

Copy link
Copy Markdown
Collaborator

PR_Github #56764 [ run ] triggered by Bot. Commit: ceb1189 Link to invocation

@tensorrt-cicd

Copy link
Copy Markdown
Collaborator

PR_Github #56764 [ run ] completed with state SUCCESS. Commit: ceb1189
/LLM/main/L0_MergeRequest_PR pipeline #45585 (Partly Tested) completed with status: 'FAILURE'

CI Report

⚠️ Action Required:

  • Please check the failed tests and fix your PR
  • If you cannot view the failures, ask the CI triggerer to share details
  • Once fixed, request an NVIDIA team member to trigger CI again

CI Agent Failure Analysis

Link to invocation

@xwang233

xwang233 commented Jul 1, 2026

Copy link
Copy Markdown
Collaborator Author

/bot run --stage-list "GB200-20_GPUs-5_Nodes-PyTorch-Disagg-PerfSanity-CTX1-NODE1-GPU4-GEN1-NODE4-GPU16-Post-Merge-*" --disable-fail-fast

@tensorrt-cicd

Copy link
Copy Markdown
Collaborator

PR_Github #56814 [ run ] triggered by Bot. Commit: 5cf29c8 Link to invocation

@tensorrt-cicd

Copy link
Copy Markdown
Collaborator

PR_Github #56814 [ run ] completed with state SUCCESS. Commit: 5cf29c8
/LLM/main/L0_MergeRequest_PR pipeline #45630 (Partly Tested) completed with status: 'SUCCESS'

CI Report

Link to invocation

@xwang233 xwang233 closed this Jul 6, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants