test: Add RHEL tutorial validation tests

[nv-rinig]

What does the PR do?

Adds qa/L0_rhel_tutorial_cpu and qa/L0_rhel_tutorial_gpu, the Step 4 checks of the RHEL/manylinux build tutorial (tutorials repo, Build_Guide/RHEL_Manylinux) as qa tests. Each test.sh writes its models, starts tritonserver from the image the tutorial built, and runs a small pytest suite against it:

• L0_rhel_tutorial_cpu: add_py (python backend) and add_onnx (onnxruntime backend) on CPU.
• L0_rhel_tutorial_gpu: add_torch (pytorch backend, TorchScript) on GPU.

The models are generated in-test, so the tests need nothing from /data/inferenceserver. They are meant to run inside the tutorial's images, where the server and backends come from build.py --target-platform=rhel, rather than in the standard QA image.

Note that previously these tests were shell scripts and lived in the tritonserver GitLab repository with CI. This change moves them into our public server repository and opportunistically converts the shell scripts into PyTest tests, still invoked through test.sh.

Checklist

• PR title reflects the change and is of format <commit_type>: <Title>
• Changes are described in the pull request.
• Related issues are referenced.
• Populated github labels field
• Added test plan and verified test passes.
• Verified that the PR passes existing CI.
• Verified copyright is correct on all changed files.
• Added succinct git squash message before merging ref.
• All template sections are filled out.
• Optional: Additional screenshots for behavior/output changes with before/after.

Commit Type:

• test

Related PRs:

• triton-inference-server/tutorials: "Reference RHEL qa tests in the RHEL tutorial's README" (link TBD)

Where should the reviewer start?

• qa/L0_rhel_tutorial_cpu/test.sh and qa/L0_rhel_tutorial_cpu/rhel_tutorial_test.py;
  the gpu directory mirrors them with one model.

Test plan:

Validated in CI: https://gitlab-master.nvidia.com/dl/dgx/tritonserver/-/pipelines/72033659

[greptile-apps]

Confidence Score: 5/5

[Low risk] Adds test cases for tutorial model serving.

The PR appears safe to merge for its intended tutorial-built images.

What we checked:

• CPU model names agree: Both model settings accept the same names, shapes, and types that the client sends. The tests check the expected output.

Summary

Adds self-contained checks for the RHEL/manylinux tutorial’s serving images.

• CPU tests check Python and ONNX addition models.
• GPU tests check a TorchScript addition model.
• Both scripts generate models, start tritonserver, run pytest, and stop the server.
• No actionable issues were established.

_{Reviews (2) · Last reviewed commit: "test: Add RHEL tutorial validation tests" · Reviewed by Greptile}