[2.8] Add warnings for missing study data mappings

[YuanTingHsieh]

What changed

• Added warning-level diagnostics when study_data.yaml is missing, empty, missing the submitted study, or has an empty dataset entry.
• Passed launcher loggers from Docker and K8s launchers into the shared study-data helper so operators see the warning in runtime logs.
• Updated Docker/K8s design docs to clarify that missing mappings still skip data mounts, but now emit warnings.
• Added focused unit coverage for the new warning paths.

Why

The 2.8 branch documents missing study-data mappings as skip-by-design behavior, but the previous implementation gave operators no visible signal when a study job launched without expected data mounts. This keeps the existing contract while making incomplete mappings diagnosable.

Validation

• git diff --check
• python3 -m compileall nvflare/app_opt/job_launcher/study_data.py nvflare/app_opt/job_launcher/docker_launcher.py nvflare/app_opt/job_launcher/k8s_launcher.py tests/unit_test/app_opt/job_launcher/study_data_test.py tests/unit_test/app_opt/job_launcher/docker_launcher_test.py tests/unit_test/app_opt/job_launcher/k8s_launcher_test.py
• Lightweight smoke check for the new warning path

Full pytest was not run locally because this environment is missing test/runtime dependencies such as pytest, yaml, and msgpack.

[Copilot]

Pull request overview

This PR improves the diagnosability of “no study data mounts” behavior in NVFlare’s Docker and K8s job launchers by emitting warning-level logs when study_data.yaml is missing/empty or lacks an entry for the requested study (while preserving the existing skip-mount behavior).

Changes:

• Added optional logger support to the shared study-data loader/resolver and emitted warnings for missing/empty mappings and missing/empty study entries.
• Wired Docker/K8s launchers to pass their launcher logger into the shared helper so warnings appear in runtime logs.
• Updated design docs and added/updated unit tests to cover the new warning paths.

Reviewed changes

Copilot reviewed 7 out of 7 changed files in this pull request and generated no comments.

Show a summary per file

File	Description
nvflare/app_opt/job_launcher/study_data.py	Adds warning diagnostics (gated by optional logger) for missing/empty study data mappings and missing/empty study entries.
nvflare/app_opt/job_launcher/docker_launcher.py	Passes the launcher logger into study-data helper calls so warnings surface in Docker runtime logs.
nvflare/app_opt/job_launcher/k8s_launcher.py	Passes the launcher logger into study-data helper calls so warnings surface in K8s runtime logs.
tests/unit_test/app_opt/job_launcher/study_data_test.py	Adds focused unit tests asserting warnings are logged when a logger is supplied.
tests/unit_test/app_opt/job_launcher/docker_launcher_test.py	Updates assertions to reflect logger propagation and verifies warning appears when study mapping is missing.
docs/design/job_launcher_and_job_handle.md	Clarifies that missing study entries skip mounts and now log warnings.
docs/design/docker_job_launcher_design.md	Clarifies that missing mappings/entries skip mounts and now log warnings.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[greptile-apps]

Greptile Summary

This PR adds warning-level diagnostics to the study-data mount path so operators see an explicit log message when study_data.yaml is missing, empty, or lacks an entry for the submitted study, while keeping the existing skip-by-design behaviour intact.

• study_data.py: Adds _log_warning helper and an optional logger parameter to load_study_data_file and resolve_study_dataset_mounts; an if study_data: guard in resolve_study_dataset_mounts prevents double-logging when the file is empty or absent (the empty-file warning is already emitted by load_study_data_file).
• docker_launcher.py / k8s_launcher.py: Thread logger=self.logger through both call sites so warnings appear in runtime logs; the K8s caching path correctly emits the missing-file warning only once on first load.
• Tests: Four new caplog unit tests cover all new warning branches; existing Docker and K8s integration tests gain caplog assertions to verify end-to-end warning propagation.

Confidence Score: 5/5

Safe to merge — changes are additive, touch only the logging path, and preserve all existing skip-by-design behaviour.

The change is purely additive: no control-flow alterations, no new failure modes, and no interface-breaking changes. The if study_data: guard correctly prevents double-logging and is documented via a docstring. All four new warning paths are covered by focused unit tests, and the Docker and K8s integration tests verify end-to-end warning propagation through caplog.

No files require special attention.

Important Files Changed

Filename	Overview
nvflare/app_opt/job_launcher/study_data.py	Adds _log_warning helper and optional logger parameter to both load_study_data_file and resolve_study_dataset_mounts; warnings are correctly gated (empty-dict guard prevents double-logging when file is empty/missing)
nvflare/app_opt/job_launcher/docker_launcher.py	Threads logger=self.logger into both study-data helpers so Docker launcher warnings surface in runtime logs; no behavioural change beyond new log output
nvflare/app_opt/job_launcher/k8s_launcher.py	Same logger threading as Docker launcher; cached study_data_pvc_dict path correctly emits the missing-file warning only on first load, with subsequent study lookups still warning when the dict is non-empty but missing the requested study
tests/unit_test/app_opt/job_launcher/study_data_test.py	Adds four focused caplog tests covering all four new warning paths (missing file, empty file, missing study entry, empty study entry); assertions are specific and correct
tests/unit_test/app_opt/job_launcher/docker_launcher_test.py	Updates call-site assertions to include logger=launcher.logger kwarg and adds a caplog assertion to test_launch_omits_data_mount_when_study_mapping_is_missing to verify the warning reaches the caller
tests/unit_test/app_opt/job_launcher/k8s_launcher_test.py	Adds a caplog.at_level assertion to the existing K8s missing-study test, mirroring the Docker-side integration coverage
docs/design/docker_job_launcher_design.md	Two doc sentences updated to note that missing mappings now log a warning; accurate and minimal
docs/design/job_launcher_and_job_handle.md	Three doc sentences updated to reflect warning logging on missing study entries; consistent with implementation

Sequence Diagram

sequenceDiagram
    participant L as DockerJobLauncher / K8sJobLauncher
    participant SD as study_data.py
    participant Log as self.logger

    L->>SD: "load_study_data_file(path, logger=self.logger)"
    alt file not found
        SD->>Log: warning("file '%s' was not found…")
        SD-->>L: "{}"
    else file empty
        SD->>Log: warning("file '%s' has no study entries…")
        SD-->>L: "{}"
    else file valid
        SD-->>L: "{study: {dataset: {...}}}"
    end

    L->>SD: "resolve_study_dataset_mounts(study_data, study, path, logger=self.logger)"
    alt study_data non-empty AND study missing
        SD->>Log: warning("file '%s' has no entry for study '%s'…")
        SD-->>L: []
    else study present but datasets empty
        SD->>Log: warning("entry for study '%s' has no datasets…")
        SD-->>L: []
    else study_data empty (already warned above)
        SD-->>L: []
    else datasets found
        SD-->>L: [StudyDatasetMount, …]
    end

Loading

_{Reviews (4): Last reviewed commit: "Merge branch '2.8' into codex/warn-missi..." | Re-trigger Greptile}

[YuanTingHsieh]

/build

[greptile-apps]

Greptile encountered an error while reviewing this PR. Please reach out to support@greptile.com for assistance.