test(telemetry): avoid requiring provider chunk intervals

A streamed response may contain one provider chunk, in which case no inter-chunk interval exists and the optional time_per_output_chunk histogram is empty. Keep the live Ollama test focused on metrics guaranteed by the request while deterministic plugin tests cover interval recording.

Assisted-by: Codex
Signed-off-by: Nigel Jones <jonesn@uk.ibm.com>

Author: planetf1

test/telemetry/test_metrics_backend.py

@@ -192,10 +192,6 @@ async def test_ollama_token_metrics_integration(
         },
     )
     ctx = ChatContext()
-    # A counting prompt reliably spans many output tokens, so the streaming
-    # branch always sees >=2 chunks (the time_per_output_chunk histogram only
-    # records inter-chunk intervals; a single-chunk reply like "Hello!" leaves
-    # it empty — observed in run 33163851256).
     ctx = ctx.add(Message(role="user", content="Count from 1 to 10 and nothing else"))
 
     model_options = {ModelOption.STREAM: True} if stream else {}
@@ -253,18 +249,6 @@ async def test_ollama_token_metrics_integration(
         assert ttfb_dp is not None, "TTFB should be recorded for streaming requests"
         assert ttfb_dp.sum > 0, "TTFB should be > 0"
 
-        # With MELLEA_GENERATION_CHUNK_EVENTS enabled, each chunk after the
-        # first records an inter-chunk interval.
-        tpoc_dp = _find_histogram_data_point(
-            metrics_data, "gen_ai.client.operation.time_per_output_chunk"
-        )
-        assert tpoc_dp is not None, (
-            "time_per_output_chunk should be recorded when chunk events are enabled"
-        )
-        assert tpoc_dp.count >= 1, (
-            "at least one inter-chunk interval should be recorded"
-        )
-
 
 @pytest.mark.asyncio
 @pytest.mark.openai