phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-10-08 04:54:38 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Merge branch 'main' into hide-non-openai-inference-apis

Browse source
This commit is contained in:
Matthew Farrellee 2025-09-26 10:48:34 -04:00
commit cb534281c8
714 changed files with 123149 additions and 54618 deletions
Download patch file Download diff file Expand all files Collapse all files

17
tests/integration/telemetry/test_openai_telemetry.py
View file

@ -49,16 +49,13 @@ def setup_openai_telemetry_data(llama_stack_client, text_model_id):
traces = llama_stack_client.telemetry.query_traces(limit=10)
if len(traces) >= 5: # 5 OpenAI completion traces
break
time.sleep(1)
time.sleep(0.1)
if len(traces) < 5:
pytest.fail(
f"Failed to create sufficient OpenAI completion telemetry data after 30s. Got {len(traces)} traces."
)
# Wait for 5 seconds to ensure traces has completed logging
time.sleep(5)
yield
@ -185,11 +182,13 @@ def test_openai_completion_creates_telemetry(llama_stack_client, text_model_id):
assert len(response.choices) > 0, "Response should have at least one choice"
# Wait for telemetry to be recorded
time.sleep(3)
# Check that we have more traces now
final_traces = llama_stack_client.telemetry.query_traces(limit=20)
final_count = len(final_traces)
start_time = time.time()
while time.time() - start_time < 30:
final_traces = llama_stack_client.telemetry.query_traces(limit=20)
final_count = len(final_traces)
if final_count > initial_count:
break
time.sleep(0.1)
# Should have at least as many traces as before (might have more due to other activity)
assert final_count >= initial_count, "Should have at least as many traces after OpenAI call"

5
tests/integration/telemetry/test_telemetry_metrics.py
View file

@ -46,10 +46,7 @@ def setup_telemetry_metrics_data(openai_client, client_with_models, text_model_i
break
except Exception:
pass
time.sleep(1)
# Wait additional time to ensure all metrics are processed
time.sleep(5)
time.sleep(0.1)
# Return the token lists for use in tests
return {"prompt_tokens": prompt_tokens, "completion_tokens": completion_tokens, "total_tokens": total_tokens}