[https://nvbugs/5448437][fix] fix some nixl tests #6940
Conversation
📝 WalkthroughWalkthroughGate two disaggregated accuracy tests to Hopper with Changes
Sequence Diagram(s)sequenceDiagram
participant Runner as PyTest Runner
participant Test as accuracy test (test_nixl_backend)
participant Conf as conftest (skip_no_hopper)
Runner->>Test: start test_nixl_backend
Test->>Conf: evaluate @skip_no_hopper
alt No Hopper
Conf-->>Runner: skip test
else Hopper
Test-->>Runner: run test body
end
sequenceDiagram
participant Runner as PyTest Runner
participant Bench as test_disaggregated_benchmark_on_diff_backends
participant Conf as defs.conftest (get_sm_version)
Runner->>Bench: start DeepSeek-V3-Lite-fp8 case
Bench->>Conf: get_sm_version()
alt SM != 90
Bench-->>Runner: skip this case
else SM == 90
Bench-->>Runner: proceed with benchmark (max_seq_len=384)
end
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~20 minutes Possibly related PRs
Suggested labels
Suggested reviewers
Tip 🔌 Remote MCP (Model Context Protocol) integration is now available!Pro plan users can now connect to remote MCP servers from the Integrations page. Connect with popular remote MCPs such as Notion and Linear to add more context to your reviews and chats. ✨ Finishing Touches
🧪 Generate unit tests
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. 🪧 TipsChatThere are 3 ways to chat with CodeRabbit:
SupportNeed help? Create a ticket on our support page for assistance with any issues or questions. CodeRabbit Commands (Invoked using PR/Issue comments)Type Other keywords and placeholders
Status, Documentation and Community
|
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 0
🔭 Outside diff range comments (1)
tests/integration/defs/disaggregated/test_disaggregated.py (1)
1189-1204: Guard process cleanup to avoid UnboundLocalError on early failures.If popen/open fails before assigning workers_proc or server_proc, the finally block will reference undefined variables.
Apply this diff to make cleanup robust:
- try: + workers_proc = None + server_proc = None + try: with ( # Start workers open('output_workers.log', 'w') as output_workers, popen(workers_cmd, stdout=output_workers, stderr=subprocess.STDOUT, env=run_env, cwd=cwd) as workers_proc, # Start server open('output_disagg.log', 'w') as output_disagg, popen(server_cmd, stdout=output_disagg, stderr=subprocess.STDOUT, env=run_env, cwd=cwd) as server_proc): # Ensure the server has started ... except Exception: ... finally: - server_proc.terminate() - workers_proc.terminate() - server_proc.wait() - workers_proc.wait() + if server_proc is not None: + server_proc.terminate() + server_proc.wait() + if workers_proc is not None: + workers_proc.terminate() + workers_proc.wait()You may want to apply the same pattern to run_disaggregated_test() for consistency.
🧹 Nitpick comments (5)
tests/integration/defs/disaggregated/test_disaggregated.py (3)
23-24: Minor: Consider consistent import style across tests.This file uses absolute imports (defs.conftest), while sibling tests use relative (..conftest). Pick one convention repo-wide to avoid path issues under different test runners.
1251-1253: Runtime gating for DeepSeek-V3-Lite FP8 on non-Hopper is sensible; consider shifting to param-level skip.Current in-test skip works; optionally, you could skip at parametrization time (e.g., using pytest.param with a mark) or apply a decorator conditioned on the model under test for clearer reporting in CI.
1130-1130: Nit: Fix log typo."Ensure the sever has started" → "Ensure the server has started".
Apply this diff:
- # Ensure the sever has started + # Ensure the server has startedtests/integration/defs/accuracy/test_disaggregated_serving.py (2)
652-654: Remove redundant no-op lines.These two lines read the same dict key without using the result and can be safely removed.
Apply this diff:
- ctx_server_config["cache_transceiver_config"] - ctx_server_config["cache_transceiver_config"]
127-133: Align env vars with selected cache transceiver backend (or drop them).launch_disaggregated_llm unconditionally sets TRTLLM_USE_UCX_KVCACHE=1 for both ctx/gen servers, while individual tests set cache_transceiver_config to "nixl" or "default". This mismatch can be confusing and may override config depending on server precedence.
Two options:
- Preferred: rely solely on the explicit cache_transceiver_config and remove env settings:
- env_ctx = os.environ.copy() - env_ctx["TRTLLM_USE_UCX_KVCACHE"] = "1" + env_ctx = os.environ.copy() ... - env_gen = os.environ.copy() - env_gen["TRTLLM_USE_UCX_KVCACHE"] = "1" + env_gen = os.environ.copy()
- Or set env vars conditionally based on the configured backend:
- env_ctx = os.environ.copy() - env_ctx["TRTLLM_USE_UCX_KVCACHE"] = "1" + env_ctx = os.environ.copy() + ctx_backend = (ctx_server_config.get("cache_transceiver_config", {}) or {}).get("backend", "default") + if ctx_backend == "nixl": + env_ctx["TRTLLM_USE_NIXL_KVCACHE"] = "1" + elif ctx_backend in ("ucx", "default"): + env_ctx["TRTLLM_USE_UCX_KVCACHE"] = "1" ... - env_gen = os.environ.copy() - env_gen["TRTLLM_USE_UCX_KVCACHE"] = "1" + env_gen = os.environ.copy() + gen_backend = (gen_server_config.get("cache_transceiver_config", {}) or {}).get("backend", "default") + if gen_backend == "nixl": + env_gen["TRTLLM_USE_NIXL_KVCACHE"] = "1" + elif gen_backend in ("ucx", "default"): + env_gen["TRTLLM_USE_UCX_KVCACHE"] = "1"Keeps the intent unambiguous and avoids conflicting signals.
Also applies to: 148-154
📜 Review details
Configuration used: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
💡 Knowledge Base configuration:
- MCP integration is disabled by default for public repositories
- Jira integration is disabled by default for public repositories
- Linear integration is disabled by default for public repositories
You can enable these sources in your CodeRabbit configuration.
📒 Files selected for processing (4)
tests/integration/defs/accuracy/test_disaggregated_serving.py(3 hunks)tests/integration/defs/disaggregated/test_disaggregated.py(4 hunks)tests/integration/test_lists/test-db/l0_dgx_b200.yml(0 hunks)tests/integration/test_lists/waives.txt(0 hunks)
💤 Files with no reviewable changes (2)
- tests/integration/test_lists/test-db/l0_dgx_b200.yml
- tests/integration/test_lists/waives.txt
🧰 Additional context used
📓 Path-based instructions (2)
**/*.py
📄 CodeRabbit Inference Engine (CODING_GUIDELINES.md)
**/*.py: Python code must target Python 3.8+
Python indentation: 4 spaces, no tabs
Maintain module namespace in imports (from package.subpackage import foo; then use foo.SomeClass())
Python file names use snake_case
Python class names use PascalCase
Python functions/methods and local variables use snake_case; variables starting with a number get k_ prefix (e.g., k_99th_percentile)
Global variables use G_ prefixed UPPER_SNAKE_CASE (e.g., G_MY_GLOBAL)
Constants use UPPER_SNAKE_CASE in Python
Avoid shadowing variables from outer scopes in Python
Initialize all externally visible members of a Python class in init
Prefer docstrings for interfaces used outside a file; comments for local code
Use Google-style docstrings for classes and functions (Sphinx-parsable)
Document attributes/variables inline with short docstrings
Avoid reflection when simple alternatives exist (e.g., prefer explicit parameters over dict(**locals()))
In try/except, catch the narrowest exceptions possible
For duck-typing with try/except, keep try body minimal and put logic in else
Files:
tests/integration/defs/disaggregated/test_disaggregated.pytests/integration/defs/accuracy/test_disaggregated_serving.py
**/*.{cpp,cxx,cc,cu,h,hpp,hxx,hh,cuh,py}
📄 CodeRabbit Inference Engine (CODING_GUIDELINES.md)
Prepend NVIDIA copyright header (current year) to all source files
Files:
tests/integration/defs/disaggregated/test_disaggregated.pytests/integration/defs/accuracy/test_disaggregated_serving.py
🧠 Learnings (2)
📚 Learning: 2025-07-28T17:06:08.621Z
Learnt from: moraxu
PR: NVIDIA/TensorRT-LLM#6303
File: tests/integration/test_lists/qa/examples_test_list.txt:494-494
Timestamp: 2025-07-28T17:06:08.621Z
Learning: In TensorRT-LLM testing, it's common to have both CLI flow tests (test_cli_flow.py) and PyTorch API tests (test_llm_api_pytorch.py) for the same model. These serve different purposes: CLI flow tests validate the traditional command-line workflow, while PyTorch API tests validate the newer LLM API backend. Both are legitimate and should coexist.
Applied to files:
tests/integration/defs/accuracy/test_disaggregated_serving.py
📚 Learning: 2025-08-06T13:58:07.506Z
Learnt from: galagam
PR: NVIDIA/TensorRT-LLM#6487
File: tests/unittest/_torch/auto_deploy/unit/singlegpu/test_ad_trtllm_bench.py:1-12
Timestamp: 2025-08-06T13:58:07.506Z
Learning: In TensorRT-LLM, test files (files under tests/ directories) do not require NVIDIA copyright headers, unlike production source code files. Test files typically start directly with imports, docstrings, or code.
Applied to files:
tests/integration/defs/accuracy/test_disaggregated_serving.py
🔇 Additional comments (5)
tests/integration/defs/disaggregated/test_disaggregated.py (2)
23-24: Imports for gating helpers look correct.Bringing in get_sm_version and skip_no_hopper matches the new gating logic introduced in this PR.
1216-1217: max_seq_len bumped to 384 aligns with benchmark settings.Given random-input-len=256 and random-output-len=64, a 384 cap provides headroom and avoids premature truncation. Looks good for both context and generation servers.
Also applies to: 1232-1233
tests/integration/defs/accuracy/test_disaggregated_serving.py (3)
24-24: Importing skip_no_hopper here is appropriate.Matches the Hopper gating applied to the nixl_backend tests.
506-506: Hopper-only gating for DeepSeekV3Lite nixl_backend is correct.Aligns with the hardware capability requirement introduced elsewhere in the PR.
638-638: Hopper-only gating for Qwen3 nixl_backend is correct.Keeps fp8/nixl paths constrained to Hopper environments.
bo-nv
force-pushed
the
user/bo/release/1.0/fix-nixl-tests
branch
from
4a09d88 to
384e7c7
Compare
bo-nv
changed the title
|
/bot run --add-multi-gpu-test |
|
PR_Github #15424 [ run ] triggered by Bot |
|
PR_Github #15424 [ run ] completed with state |
bo-nv
force-pushed
the
user/bo/release/1.0/fix-nixl-tests
branch
from
9743e5f to
eafcfab
Compare
|
/bot run --add-multi-gpu-test |
|
PR_Github #15529 [ run ] triggered by Bot |
|
PR_Github #15529 [ run ] completed with state |
|
/bot run --add-multi-gpu-test |
|
PR_Github #15540 [ run ] triggered by Bot |
|
PR_Github #15540 [ run ] completed with state |
|
/bot run --add-multi-gpu-test |
|
PR_Github #15553 [ run ] triggered by Bot |
bo-nv
force-pushed
the
user/bo/release/1.0/fix-nixl-tests
branch
from
eafcfab to
5dd8e94
Compare
|
/bot run --add-multi-gpu-test |
|
PR_Github #15635 [ run ] triggered by Bot |
|
PR_Github #15553 [ run ] completed with state |
|
PR_Github #15635 [ run ] completed with state |
Signed-off-by: Bo Deng <[email protected]>
Signed-off-by: Bo Deng <[email protected]>
bo-nv
force-pushed
the
user/bo/release/1.0/fix-nixl-tests
branch
from
5dd8e94 to
5caa636
Compare
|
/bot run --add-multi-gpu-test |
There was a problem hiding this comment.
Actionable comments posted: 0
🧹 Nitpick comments (2)
tests/integration/defs/disaggregated/test_disaggregated.py (2)
1216-1217: max_seq_len bumped to 384 matches benchmark input/output lengthsThe new 384 setting provides headroom over 256+64=320 used by the benchmark and is consistent across context and generation servers.
Consider factoring 384 into a named constant (e.g., MAX_CONTEXT_LEN) or deriving it from the benchmark parameters to avoid drift between the client args and server config.
Also applies to: 1232-1233
1251-1253: Nit: simplify get_sm_version() call formatting and clarify skip messageSingle-line the call for readability and make the reason explicit.
- if "DeepSeek-V3-Lite" in benchmark_model_root and "fp8" in benchmark_model_root and get_sm_version( - ) != 90: - pytest.skip("The test should only run on Hopper") + if ("DeepSeek-V3-Lite" in benchmark_model_root + and "fp8" in benchmark_model_root + and get_sm_version() != 90): + pytest.skip("Gated: DeepSeek-V3-Lite FP8 is only supported on Hopper (SM90)")
📜 Review details
Configuration used: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
💡 Knowledge Base configuration:
- MCP integration is disabled by default for public repositories
- Jira integration is disabled by default for public repositories
- Linear integration is disabled by default for public repositories
You can enable these sources in your CodeRabbit configuration.
📒 Files selected for processing (4)
tests/integration/defs/accuracy/test_disaggregated_serving.py(3 hunks)tests/integration/defs/disaggregated/test_disaggregated.py(4 hunks)tests/integration/test_lists/test-db/l0_dgx_b200.yml(0 hunks)tests/integration/test_lists/waives.txt(0 hunks)
💤 Files with no reviewable changes (2)
- tests/integration/test_lists/test-db/l0_dgx_b200.yml
- tests/integration/test_lists/waives.txt
🚧 Files skipped from review as they are similar to previous changes (1)
- tests/integration/defs/accuracy/test_disaggregated_serving.py
🧰 Additional context used
📓 Path-based instructions (2)
**/*.py
📄 CodeRabbit Inference Engine (CODING_GUIDELINES.md)
**/*.py: Python code must target Python 3.8+
Python indentation: 4 spaces, no tabs
Maintain module namespace in imports (from package.subpackage import foo; then use foo.SomeClass())
Python file names use snake_case
Python class names use PascalCase
Python functions/methods and local variables use snake_case; variables starting with a number get k_ prefix (e.g., k_99th_percentile)
Global variables use G_ prefixed UPPER_SNAKE_CASE (e.g., G_MY_GLOBAL)
Constants use UPPER_SNAKE_CASE in Python
Avoid shadowing variables from outer scopes in Python
Initialize all externally visible members of a Python class in init
Prefer docstrings for interfaces used outside a file; comments for local code
Use Google-style docstrings for classes and functions (Sphinx-parsable)
Document attributes/variables inline with short docstrings
Avoid reflection when simple alternatives exist (e.g., prefer explicit parameters over dict(**locals()))
In try/except, catch the narrowest exceptions possible
For duck-typing with try/except, keep try body minimal and put logic in else
Files:
tests/integration/defs/disaggregated/test_disaggregated.py
**/*.{cpp,cxx,cc,cu,h,hpp,hxx,hh,cuh,py}
📄 CodeRabbit Inference Engine (CODING_GUIDELINES.md)
Prepend NVIDIA copyright header (current year) to all source files
Files:
tests/integration/defs/disaggregated/test_disaggregated.py
⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (1)
- GitHub Check: Pre-commit Check
🔇 Additional comments (1)
tests/integration/defs/disaggregated/test_disaggregated.py (1)
23-24: LGTM: Importing get_sm_version and skip_no_hopperThe extended import list aligns with the new runtime gating below and keeps related skips centralized in conftest.
|
PR_Github #15718 [ run ] triggered by Bot |
|
PR_Github #15718 [ run ] completed with state |
|
/bot run --add-multi-gpu-test |
1 similar comment
|
/bot run --add-multi-gpu-test |
|
PR_Github #15733 [ run ] triggered by Bot |
|
PR_Github #15733 [ run ] completed with state |
|
/bot run --add-multi-gpu-test |
|
PR_Github #15837 [ run ] triggered by Bot |
bo-nv
changed the title
bo-nv
changed the title
|
PR_Github #15837 [ run ] completed with state |
Signed-off-by: Bo Deng <[email protected]>
Signed-off-by: Bo Deng <[email protected]>
Signed-off-by: Bo Deng <[email protected]>
Signed-off-by: Bo Deng <[email protected]>
Signed-off-by: Bo Deng <[email protected]>
Signed-off-by: Bo Deng <[email protected]>
Summary by CodeRabbit
This MR will fix:
https://nvbugs/5448437
https://nvbugs/5448437
Description
Test Coverage
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.
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)]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. Examples: "A10-PyTorch-1, xxx". 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. Examples: --extra-stage "H100_PCIe-TensorRT-Post-Merge-1, xxx".--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 thestage-listparameter to access the appropriate container environment. Note: Does NOT update GitHub check status.For guidance on mapping tests to stage names, see
docs/source/reference/ci-overview.mdand the
scripts/test_to_stage_mapping.pyhelper.kill
killKill all running builds associated with pull request.
skip
skip --comment COMMENTSkip 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-pipelineReuse 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.