test(telemetry): avoid requiring provider chunk intervals

[planetf1]

Pull Request

Issue

Partly addresses #1566 (test case impact on other code)

Description

The live Ollama telemetry test required gen_ai.client.operation.time_per_output_chunk to exist for every streamed response. That metric records inter-chunk intervals, so a valid one-chunk response has no interval and leaves the histogram empty. Provider chunk framing is not deterministic, even when the prompt produces several output tokens, so the assertion caused unrelated CI failures.

Remove that provider-dependent assertion while retaining the live checks for token usage, request duration, and time to first chunk. Deterministic unit coverage for interval recording remains unchanged. The broader producer-side telemetry correction stays scoped to issue #1566.

Testing

• Tests added to the respective file if code was changed
• New code has 100% coverage if code was added
• Ensure existing tests and github automation passes (a maintainer will kick off the github automation when the rest of the PR is populated)

Verified locally:

uv run pytest test/telemetry/test_metrics_plugins.py test/helpers/test_async_helpers.py -q
18 passed, 1 skipped
uv run ruff check test/telemetry/test_metrics_backend.py
All checks passed!
uv run ruff format --check test/telemetry/test_metrics_backend.py
1 file already formatted

Attribution

• AI coding assistants used

---

Adding a new component, requirement, sampling strategy, or tool?

If your PR adds or modifies one of the types below, check the matching box. A checklist of type-specific review items will be posted as a comment.

• Component
• Requirement
• Sampling Strategy
• Tool

NOTE: Please ensure you have an issue that has been acknowledged by a core contributor and routed you to open a pull request against this repository. Otherwise, please open an issue before continuing with this pull request.

[AngeloDanducci]

LGTM

[ajbozarth]

I'm ok with removing this test assert if you've found it flaky.