[samsullivan1]

Purpose

Wire the three vLLM framework test suites — vllm/upstream, vllm/model, vllm/sagemaker — into the content-addressed test-skip cache and enable them end-to-end. CI can now skip a vLLM suite when a prior PASS exists for the same image content + test code, and it records a PASS after the suite genuinely ran and passed. The mechanism is fail-open: any cache miss, degraded lookup, or infra failure runs the suite as normal.

• .github/workflows/vllm.tests-upstream.yml, vllm.tests-model.yml, vllm.tests-sagemaker.yml — add optional image-content-hash / suite-code-hash inputs and a record-test-pass job to each reusable. A PASS is recorded only when the suite genuinely ran and passed: the record job requires the test jobs to report success (upstream: all of regression/cuda/example; model: at least one non-empty matrix job, with none failed; sagemaker: endpoint-test). Because each test job is already gated at the job level on the preflight image probe, an absent image (or empty matrix) leaves those jobs skipped — never success — so a self-skipped run never records a false PASS. Shared by all callers; a no-op for callers that pass no hashes.
• .github/workflows/vllm.pipeline.yml — add the three vLLM suites to the pipeline's check job that looks up prior passes, gate upstream-tests / model-tests / sagemaker-tests on the result (fail-open via skips || '{}', gated only on build failure — never on the read-only check job's status), and thread the computed hashes into each reusable so a pass is recorded. Existing gates (e.g. sagemaker-tests requiring customer-type == 'sagemaker') are preserved.
• .github/workflows/huggingface-vllm.pipeline.yml — this pipeline reuses vllm.tests-sagemaker.yml, so vllm/sagemaker is added to its check job and its sagemaker-tests caller is wired the same way.
• .github/config/test-suites.yml — extend the code_paths for the matrix-driven suites so a change to the tested-model set invalidates the cache alongside test/vllm/**: vllm/model gains .github/config/model-tests/vllm-model-tests.yml and vllm/sagemaker gains .github/config/model-tests/vllm-sagemaker-endpoint-tests.yml.

Pipelines wired: vllm, huggingface-vllm (sagemaker suite only).

The check and record-test-pass jobs are both fail-open, so a degraded cache never blocks CI; on a cold cache the suite simply runs and populates a row on pass.

Test Plan

• python -m pytest test/db — reconciliation tests (invoked suites exist in config; skip accessors are configured and checked) plus the suite/image hashing unit tests.
• Verified all edited workflow files parse as YAML.
• Confirmed the reconciliation checks detect the three vllm/* accessors in each wired pipeline's job if: conditions AND that each accessor is present in that pipeline's check-job suites list.

Test Result

• test/db: 37 passed.
• All edited *.yml parse as YAML; the vllm/upstream, vllm/model, and vllm/sagemaker accessors are detected in the check-job suites list and in the gated jobs' if: conditions for vllm.pipeline.yml, and vllm/sagemaker likewise for huggingface-vllm.pipeline.yml.

Live CI validation

Validated via test PR #6627. Record→hit is validated without a live endpoint: vllm/sagemaker can't run in PR CI due to recurring GPU InsufficientInstanceCapacity (the endpoint never reaches InService, so the suite never passes and correctly records nothing — fail-open). The record/hit logic is suite-agnostic shared code, covered by moto unit tests in test/db (Database Unit Tests workflow), and its live record→hit was proven on capacity-independent suites — telemetry (#6599), EFA (#6602), Ray (#6633), PyTorch (#6632) — each recording on run 1 and getting a cache HIT (suite skipped) on a same-day re-run.

---

Toggle if you are merging into master Branch

By default, docker image builds and tests are disabled. Two ways to run builds and tests:

1. Using dlc_developer_config.toml
2. Using this PR description (currently only supported for PyTorch, TensorFlow, vllm, and base images)

How to use the helper utility for updating dlc_developer_config.toml

Assuming your remote is called origin (you can find out more with git remote -v)...

• Run default builds and tests for a particular buildspec - also commits and pushes changes to remote; Example:

python src/prepare_dlc_dev_environment.py -b </path/to/buildspec.yml> -cp origin

• Enable specific tests for a buildspec or set of buildspecs - also commits and pushes changes to remote; Example:

python src/prepare_dlc_dev_environment.py -b </path/to/buildspec.yml> -t sanity_tests -cp origin

• Restore TOML file when ready to merge

python src/prepare_dlc_dev_environment.py -rcp origin

NOTE: If you are creating a PR for a new framework version, please ensure success of the local, standard, rc, and efa sagemaker tests by updating the dlc_developer_config.toml file:

• sagemaker_remote_tests = true
• sagemaker_efa_tests = true
• sagemaker_rc_tests = true
• sagemaker_local_tests = true

How to use PR description

Use the code block below to uncomment commands and run the PR CodeBuild jobs. There are two commands available:

• # /buildspec <buildspec_path>
  • e.g.: # /buildspec pytorch/training/buildspec.yml
  • If this line is commented out, dlc_developer_config.toml will be used.
• # /tests <test_list>
  • e.g.: # /tests sanity security ec2
  • If this line is commented out, it will run the default set of tests (same as the defaults in dlc_developer_config.toml): sanity, security, ec2, ecs, eks, sagemaker, sagemaker-local.

# /buildspec <buildspec_path>
# /tests <test_list>

Toggle if you are merging into main Branch

PR Checklist

• I ran pre-commit run --all-files locally before creating this PR. (Read DEVELOPMENT.md for details).