[https://nvbugs/6314696] [fix] update disaggregated GB200 test configs by reasonsolo · Pull Request #16481 · NVIDIA/TensorRT-LLM

reasonsolo · 2026-07-16T07:07:41Z

~~The gen-only tests were removed.~~
disagg-e2e-gb200_deepseek-r1-fp4_128k8k_con64_ctx1_pp8_gen1_dep32_eplb0_mtp3_ccb-NIXL test case itself doesn't make sense. Long ISL/OSL, high concurrency, but small max batch size 1, led to long running time (>2H). Increasing the MBS led to OOM.
disagg-e2e-gb200_deepseek-v32-fp4_32k4k_con256_ctx1_dep8_gen1_dep8_eplb0_mtp0_ccb-NIXL has batch_size* DEP < concurrency, which will be rejected by ADP overlap executor loop.

Summary by CodeRabbit

Performance Testing
- Updated GB200 GPU-conditioned performance test coverage and configurations.
- Removed several obsolete DeepSeek-R1, DeepSeek-V32, WideEP, and Qwen3 benchmark scenarios.
- Increased the DeepSeek-V32 generation maximum batch size from 8 to 32.
- Removed an outdated DeepSeek-R1 benchmark configuration.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.

Signed-off-by: Lizhi Zhou <1432185+reasonsolo@users.noreply.github.com>

coderabbitai · 2026-07-16T07:10:47Z

📝 Walkthrough

Walkthrough

GB200 disaggregated performance test coverage was revised, including removal and replacement of selected DeepSeek and Qwen configurations. The DeepSeek-V32 generation batch-size limits were increased from 8 to 32. The GB300 section was unchanged.

Changes

GB200 performance updates

Layer / File(s)	Summary
DeepSeek-V32 batch-size configuration `tests/scripts/perf-sanity/disaggregated/gb200_deepseek-v32-...yaml`	Generation and CUDA graph maximum batch sizes were increased from 8 to 32.
Disaggregated test matrix alignment `tests/integration/test_lists/qa/llm_perf_disagg.yml`, `tests/integration/test_lists/qa/llm_perf_multinode.txt`, `tests/scripts/perf-sanity/disaggregated/gb200_deepseek-r1-...yaml`	Selected GB200 DeepSeek-R1, DeepSeek-V32, and Qwen3-235B configurations were removed or replaced in disaggregated performance lists; the DeepSeek-R1 benchmark configuration was deleted.

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

Possibly related PRs

NVIDIA/TensorRT-LLM#16474: Updates waiver entries for overlapping disaggregated generation-only performance combinations.

Suggested reviewers: zhenhuaw-me

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Description check	⚠️ Warning	The description lacks the required Description and Test Coverage sections and does not follow the repository template.	Add the Description and Test Coverage headings, briefly explain the issue and fix, and keep the checklist item completed if applicable.

✅ Passed checks (4 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 follows the required ticket/type format and clearly summarizes the GB200 disaggregated test config update.

✨ Finishing Touches

🧪 Generate unit tests (beta)

Create PR with unit tests

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

nv-xtf

LGTM

Signed-off-by: Lizhi Zhou <1432185+reasonsolo@users.noreply.github.com>

reasonsolo · 2026-07-16T07:45:17Z

/bot run --disable-fail-fast

tensorrt-cicd · 2026-07-16T07:52:46Z

PR_Github #59663 [ run ] triggered by Bot. Commit: d85782c Link to invocation

BowenFu

LGTM — GB200 disagg config cleanup: removes the >2h/OOM con64 case and fixes v32 con256_dep8 max_batch_size (256≥concurrency); stage/yml/waive updated consistently.

fredricz-20070104 · 2026-07-16T12:01:34Z

QA disagg perf on GB200 runs off qa/llm_perf_multinode.txt, not qa/llm_perf_disagg.yml (bug log path is .../tests_mulinode/).

This PR only removed the cases from llm_perf_disagg.yml, but they're still in qa/llm_perf_multinode.txt. If you want to remove them, please also remove them from qa/llm_perf_multinode.txt — otherwise the QA path keeps failing and the bug isn't fixed.

tensorrt-cicd · 2026-07-16T21:40:07Z

PR_Github #59663 [ run ] completed with state SUCCESS. Commit: d85782c
/LLM/main/L0_MergeRequest_PR pipeline #48097 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

Signed-off-by: Lizhi Zhou <1432185+reasonsolo@users.noreply.github.com>

fredricz-20070104

Approved.

Signed-off-by: Lizhi Zhou <1432185+reasonsolo@users.noreply.github.com>

reasonsolo · 2026-07-17T05:24:30Z

/bot run --disable-fail-fast

tensorrt-cicd · 2026-07-17T05:30:50Z

PR_Github #59887 [ run ] triggered by Bot. Commit: 3c4244b Link to invocation

reasonsolo · 2026-07-17T05:39:32Z

/bot help

github-actions · 2026-07-17T05:39:40Z

GitHub Bot Help

/bot [-h] ['run', 'kill', 'skip', 'reuse-pipeline'] ...

Provide a user friendly way for developers to interact with a Jenkins server.

Run /bot [-h|--help] to print this help message.

See details below for each supported subcommand.

Details

run [--reuse-test (optional)pipeline-id --disable-fail-fast --skip-test --stage-list "A10-PyTorch-1, xxx" --gpu-type "A30, H100_PCIe" --test-backend "pytorch, cpp" --add-multi-gpu-test --only-multi-gpu-test --disable-multi-gpu-test --post-merge --extra-stage "H100_PCIe-TensorRT-Post-Merge-1, xxx" --detailed-log --debug(experimental) --high-priority]

Launch build/test pipelines. All previously running jobs will be killed.

--reuse-test (optional)pipeline-id (OPTIONAL) : Allow the new pipeline to reuse build artifacts and skip successful test stages from a specified pipeline or the last pipeline if no pipeline-id is indicated. If the Git commit ID has changed, this option will be always ignored. The DEFAULT behavior of the bot is to reuse build artifacts and successful test results from the last pipeline.

--disable-reuse-test (OPTIONAL) : Explicitly prevent the pipeline from reusing build artifacts and skipping successful test stages from a previous pipeline. Ensure that all builds and tests are run regardless of previous successes.

--disable-fail-fast (OPTIONAL) : Disable fail fast on build/tests/infra failures.

--skip-test (OPTIONAL) : Skip all test stages, but still run build stages, package stages and sanity check stages. Note: Does NOT update GitHub check status.

--stage-list "A10-PyTorch-1, xxx" (OPTIONAL) : Only run the specified test stages. Supports wildcard * for pattern matching (e.g., "*PerfSanity*" matches all stages containing PerfSanity). Examples: "A10-PyTorch-1, xxx", "PerfSanity". Note: Does NOT update GitHub check status.

--gpu-type "A30, H100_PCIe" (OPTIONAL) : Only run the test stages on the specified GPU types. Examples: "A30, H100_PCIe". Note: Does NOT update GitHub check status.

--test-backend "pytorch, cpp" (OPTIONAL) : Skip test stages which don't match the specified backends. Only support [pytorch, cpp, tensorrt, triton]. Examples: "pytorch, cpp" (does not run test stages with tensorrt or triton backend). Note: Does NOT update GitHub pipeline status.

--only-multi-gpu-test (OPTIONAL) : Only run the multi-GPU tests. Note: Does NOT update GitHub check status.

--disable-multi-gpu-test (OPTIONAL) : Disable the multi-GPU tests. Note: Does NOT update GitHub check status.

--add-multi-gpu-test (OPTIONAL) : Force run the multi-GPU tests in addition to running L0 pre-merge pipeline.

--post-merge (OPTIONAL) : Run the L0 post-merge pipeline instead of the ordinary L0 pre-merge pipeline.

--extra-stage "H100_PCIe-TensorRT-Post-Merge-1, xxx" (OPTIONAL) : Run the ordinary L0 pre-merge pipeline and specified test stages. Supports wildcard * for pattern matching. Examples: --extra-stage "H100_PCIe-TensorRT-Post-Merge-1, xxx", --extra-stage "Post-Merge".

--detailed-log (OPTIONAL) : Enable flushing out all logs to the Jenkins console. This will significantly increase the log volume and may slow down the job.

--debug (OPTIONAL) : Experimental feature. Enable access to the CI container for debugging purpose. Note: Specify exactly one stage in the stage-list parameter to access the appropriate container environment. Note: Does NOT update GitHub check status.

--high-priority (OPTIONAL) : Run the pipeline with high priority. This option is restricted to authorized users only and will route the job to a high-priority queue.

kill

kill

Kill all running builds associated with pull request.

skip

skip --comment COMMENT

Skip testing for latest commit on pull request. --comment "Reason for skipping build/test" is required. IMPORTANT NOTE: This is dangerous since lack of user care and validation can cause top of tree to break.

reuse-pipeline

reuse-pipeline

Reuse a previous pipeline to validate current commit. This action will also kill all currently running builds associated with the pull request. IMPORTANT NOTE: This is dangerous since lack of user care and validation can cause top of tree to break.

reasonsolo · 2026-07-17T05:40:57Z

/bot run --stage-list "GB200-16_GPUs-4_Nodes-PyTorch-Disagg-PerfSanity-CTX1-NODE2-GPU8-GEN1-NODE2-GPU8-Post-Merge"

tensorrt-cicd · 2026-07-17T05:46:32Z

PR_Github #59895 [ run ] triggered by Bot. Commit: 3c4244b Link to invocation

tensorrt-cicd · 2026-07-17T05:50:09Z

PR_Github #59887 [ run ] completed with state ABORTED. Commit: 3c4244b

Link to invocation

tensorrt-cicd · 2026-07-17T06:06:21Z

PR_Github #59895 [ run ] completed with state FAILURE. Commit: 3c4244b
/LLM/main/L0_MergeRequest_PR pipeline #48297 (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

reasonsolo · 2026-07-17T06:16:50Z

/bot run --disable-fail-fast

tensorrt-cicd · 2026-07-17T06:22:19Z

PR_Github #59902 [ run ] triggered by Bot. Commit: 3c4244b Link to invocation

tensorrt-cicd · 2026-07-17T09:37:15Z

PR_Github #59902 [ run ] completed with state SUCCESS. Commit: 3c4244b
/LLM/main/L0_MergeRequest_PR pipeline #48306 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

reasonsolo · 2026-07-17T09:56:26Z

/bot run --disable-fail-fast

tensorrt-cicd · 2026-07-17T10:02:19Z

PR_Github #59957 [ run ] triggered by Bot. Commit: 3c4244b Link to invocation

tensorrt-cicd · 2026-07-17T14:02:19Z

PR_Github #59957 [ run ] completed with state FAILURE. Commit: 3c4244b
/LLM/main/L0_MergeRequest_PR pipeline #48354 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

reasonsolo · 2026-07-17T14:35:55Z

/bot run --disable-fail-fast

tensorrt-cicd · 2026-07-17T14:41:26Z

PR_Github #59993 [ run ] triggered by Bot. Commit: 3c4244b Link to invocation

fix(perf): update disaggregated GB200 test configs

1c9dced

Signed-off-by: Lizhi Zhou <1432185+reasonsolo@users.noreply.github.com>

reasonsolo requested review from a team as code owners July 16, 2026 07:07

reasonsolo requested review from Tabrizian, chenfeiz0326, dhansen-nvidia, nv-xtf, ruodil and yingguo-trt July 16, 2026 07:07

github-actions Bot assigned reasonsolo Jul 16, 2026

yingguo-trt approved these changes Jul 16, 2026

View reviewed changes

nv-xtf approved these changes Jul 16, 2026

View reviewed changes

fix(ci): remove obsolete GB200 test stage

d85782c

Signed-off-by: Lizhi Zhou <1432185+reasonsolo@users.noreply.github.com>

reasonsolo requested review from a team as code owners July 16, 2026 07:34

reasonsolo requested review from BowenFu, brnguyen2, joyang-nv, longlee0622, yiqingy0 and yuanjingx87 July 16, 2026 07:34

yufeiwu-nv removed the request for review from ruodil July 16, 2026 07:36

chenfeiz0326 approved these changes Jul 16, 2026

View reviewed changes

BowenFu approved these changes Jul 16, 2026

View reviewed changes

brnguyen2 approved these changes Jul 16, 2026

View reviewed changes

fix(perf): remove disabled multinode test cases

c682318

Signed-off-by: Lizhi Zhou <1432185+reasonsolo@users.noreply.github.com>

fredricz-20070104 approved these changes Jul 17, 2026

View reviewed changes

reasonsolo added 2 commits July 16, 2026 22:05

fix(perf): restore verified GB200 gen-only tests

6a2504b

Signed-off-by: Lizhi Zhou <1432185+reasonsolo@users.noreply.github.com>

Merge branch 'main' into bug6314696

3c4244b

Signed-off-by: Lizhi Zhou <1432185+reasonsolo@users.noreply.github.com>

reasonsolo enabled auto-merge (squash) July 17, 2026 05:24

longlee0622 approved these changes Jul 17, 2026

View reviewed changes

Uh oh!

Conversation

reasonsolo commented Jul 16, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary by CodeRabbit

Description

Test Coverage

PR Checklist

GitHub Bot Help

Uh oh!

coderabbitai Bot commented Jul 16, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Walkthrough

Changes

❌ Failed checks (1 warning)

Uh oh!

nv-xtf left a comment

Choose a reason for hiding this comment

Uh oh!

reasonsolo commented Jul 16, 2026

Uh oh!

tensorrt-cicd commented Jul 16, 2026

Uh oh!

BowenFu left a comment

Choose a reason for hiding this comment

Uh oh!

fredricz-20070104 commented Jul 16, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

tensorrt-cicd commented Jul 16, 2026

Uh oh!

fredricz-20070104 left a comment

Choose a reason for hiding this comment

Uh oh!

reasonsolo commented Jul 17, 2026

Uh oh!

tensorrt-cicd commented Jul 17, 2026

Uh oh!

reasonsolo commented Jul 17, 2026

Uh oh!

github-actions Bot commented Jul 17, 2026

GitHub Bot Help

kill

skip

reuse-pipeline

Uh oh!

reasonsolo commented Jul 17, 2026

Uh oh!

tensorrt-cicd commented Jul 17, 2026

Uh oh!

tensorrt-cicd commented Jul 17, 2026

Uh oh!

tensorrt-cicd commented Jul 17, 2026

Uh oh!

reasonsolo commented Jul 17, 2026

Uh oh!

tensorrt-cicd commented Jul 17, 2026

Uh oh!

tensorrt-cicd commented Jul 17, 2026

Uh oh!

reasonsolo commented Jul 17, 2026

Uh oh!

tensorrt-cicd commented Jul 17, 2026

Uh oh!

tensorrt-cicd commented Jul 17, 2026

Uh oh!

reasonsolo commented Jul 17, 2026

Uh oh!

tensorrt-cicd commented Jul 17, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

9 participants

reasonsolo commented Jul 16, 2026 •

edited

Loading

coderabbitai Bot commented Jul 16, 2026 •

edited

Loading

fredricz-20070104 commented Jul 16, 2026 •

edited

Loading