test(tracing): add round-trip tests for ag.metrics.duration.cumulativ…

[justin212407]

Summary:

Fixes #4172 - ag.metrics.duration.cumulative query↔ingest schema asymmetry that caused spans to be silently dropped on re-ingest.

During ingestion, parsing.py computed wall-clock duration and wrote it as a raw scalar float. The original Pydantic model (AgMetricEntryAttributes) declared cumulative as Optional[Metrics] where Metrics = Dict[str, NumericJson] - a dict, not a scalar. This meant any span queried via /api/tracing/spans/query and re-submitted to /api/tracing/spans/ingest failed Pydantic validation silently, returning {"count": 0, "links": []} with no error surfaced to the caller.

Per the team's documented decision in docs/design/homomorphic-tracing-io/findings.md, the canonical shape for duration.cumulative is a scalar float. The production fixes (AgScalarMetricEntryAttributes in tracing.py, _coerce_scalar_metric_value in attributes.py, scalar write in parsing.py) were already present in the codebase. This PR adds the missing test coverage for those changes.

Testing:

Verified locally

# New tests only
uv run pytest oss/tests/pytest/unit/tracing/utils/test_parsing.py \
  -v -k "duration_cumulative" --no-header
# 2 passed

# Full tracing suite
uv run pytest oss/tests/pytest/unit/tracing/ -v --no-header -p no:warnings
# 74 passed

# Full unit suite
uv run pytest oss/tests/pytest/unit/ -q --no-header -p no:warnings
# 525 passed

# Linting and formatting
uv run ruff check oss/src/core/tracing/ oss/tests/pytest/unit/tracing/
uv run ruff format --check oss/tests/pytest/unit/tracing/utils/test_parsing.py
# All checks passed

Added or updated tests:

Two new unit tests added to

1. api/oss/tests/pytest/unit/tracing/utils/test_parsing.py:
   test_parse_spans_from_request_accepts_scalar_duration_cumulative_through_ingest - Verifies that a span with metrics.duration.cumulative set as a scalar float passes through the full parse_spans_from_request pipeline without error and the output value is a plain float, not a dict. Covers the canonical ingest path.

2. test_parse_spans_from_request_normalizes_legacy_dict_duration_cumulative_to_scalar - Verifies backward compatibility: a span carrying the old dict shape duration.cumulative = {"total": 1922.653} (as written by older ingestion code
   or replayed from a query response) is normalized to the canonical scalar 1922.653 by _coerce_scalar_metric_value inside initialize_ag_attributes. No timestamps are supplied so the computed-duration override path is skipped,
   isolating the normalization behavior.

QA follow-up:

• Verify that existing traces in production with duration.cumulative as a scalar are still returned correctly by /api/tracing/spans/query.
• Verify that a full query→re-ingest round-trip (the exact repro from the issue) no longer returns {"count": 0, "links": []}.
• Verify that traces ingested by older SDK versions (which may have written duration.cumulative = {"total": v}) are normalized correctly on the way through ingest without requiring a backfill.

Demo:

N/A - no UI changes.

Checklist:

• I have included a video or screen recording for UI changes, or marked Demo as N/A.
• Relevant tests pass locally.
• Relevant linting and formatting pass locally.
• I have signed the CLA, or I will sign it when the bot prompts me.

Contributor Resources:

• Contributing guide
• Creating your first PR
• Development mode
• Testing
• Formatting and linting
• Slack support

[vercel]

@justin212407 is attempting to deploy a commit to the agenta projects Team on Vercel.

A member of the Team first needs to authorize it.

[coderabbitai]

📝 Walkthrough

Walkthrough

Two pytest unit tests are added to test_parsing.py to validate parse_spans_from_request metric duration normalization. The first test verifies that scalar ag.metrics.duration.cumulative values are accepted and overwritten by computed wall-clock duration when start_time and end_time are present. The second test confirms that legacy dict-shaped duration metrics ({"total": ...}) are normalized into scalar form when timestamps are absent.

Changes

Metric duration ingestion normalization tests

Layer / File(s)	Summary
Scalar and dict metric duration handling api/oss/tests/pytest/unit/tracing/utils/test_parsing.py	parse_spans_from_request accepts canonical scalar ag.metrics.duration.cumulative values and overwrites them with computed wall-clock duration when timestamps are present, keeping the result scalar; it also normalizes legacy dict shape ({"total": ...}) into scalar when timestamps are absent.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~8 minutes

🚥 Pre-merge checks | ✅ 4 | ❌ 1

❌ Failed checks (1 warning)

Check name	Status	Explanation	Resolution
Docstring Coverage	⚠️ Warning	Docstring coverage is 0.00% which is insufficient. The required threshold is 60.00%.	Write docstrings for the functions missing them to satisfy the coverage threshold.

✅ Passed checks (4 passed)

Check name	Status	Explanation
Title check	✅ Passed	The PR title correctly identifies the main change: adding test coverage for ag.metrics.duration.cumulative round-trip behavior.
Linked Issues check	✅ Passed	The PR adds test coverage for the production fixes that resolve issue #4172's schema asymmetry: it verifies scalar duration.cumulative acceptance through ingest and legacy dict-to-scalar normalization.
Out of Scope Changes check	✅ Passed	The PR contains only test additions to verify existing production code behavior; no out-of-scope changes to production code are present.
Description check	✅ Passed	The pull request description is clearly related to the changeset. It explains the issue being fixed, the production code changes already in place, and documents the test coverage being added for schema asymmetry in ag.metrics.duration.cumulative.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with 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.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

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

[Copilot]

Pull request overview

Note

Copilot was unable to run its full agentic suite in this review.

Adds unit tests around parse_spans_from_request to ensure duration metric normalization and precedence rules are stable across ingest/parsing.

Changes:

• Adds a test ensuring scalar duration.cumulative is accepted and overridden by computed duration when timestamps are present.
• Adds a test ensuring legacy {"total": v} dict shape for duration.cumulative is normalized to a scalar when timestamps are absent.

---

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