phoenix-oss/llama-stack-mirror
0
0
Fork 1
mirror of https://github.com/meta-llama/llama-stack.git synced 2025-12-03 18:00:36 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Merge branch 'main' into elasticsearch-integration

Browse source
This commit is contained in:
Francisco Javier Arceo 2025-12-01 16:54:45 -05:00 • committed by GitHub
commit 7aaab870bd
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
87 changed files with 762 additions and 5192 deletions
Download patch file Download diff file Expand all files Collapse all files

7
tests/integration/inference/test_provider_data_routing.py
View file

@ -17,7 +17,6 @@ from unittest.mock import AsyncMock, patch
import pytest
from llama_stack.core.library_client import LlamaStackAsLibraryClient
from llama_stack.core.telemetry.telemetry import MetricEvent
from llama_stack_api import (
Api,
OpenAIAssistantMessageParam,
@ -27,10 +26,6 @@ from llama_stack_api import (
)
class OpenAIChatCompletionWithMetrics(OpenAIChatCompletion):
metrics: list[MetricEvent] | None = None
def test_unregistered_model_routing_with_provider_data(client_with_models):
"""
Test that a model can be routed using provider_id/model_id format
@ -72,7 +67,7 @@ def test_unregistered_model_routing_with_provider_data(client_with_models):
# The inference router's routing_table.impls_by_provider_id should have anthropic
# Let's patch the anthropic provider's openai_chat_completion method
# to avoid making real API calls
mock_response = OpenAIChatCompletionWithMetrics(
mock_response = OpenAIChatCompletion(
id="chatcmpl-test-123",
created=1234567890,
model="claude-3-5-sonnet-20241022",

6
tests/integration/telemetry/collectors/in_memory.py
View file

@ -15,11 +15,10 @@ from opentelemetry.sdk.trace import TracerProvider
from opentelemetry.sdk.trace.export import SimpleSpanProcessor
from opentelemetry.sdk.trace.export.in_memory_span_exporter import InMemorySpanExporter
import llama_stack.core.telemetry.telemetry as telemetry_module
from .base import BaseTelemetryCollector, MetricStub, SpanStub
# TODO: Fix thi to work with Automatic Instrumentation
class InMemoryTelemetryCollector(BaseTelemetryCollector):
"""In-memory telemetry collector for library-client tests.
@ -75,13 +74,10 @@ class InMemoryTelemetryManager:
meter_provider = MeterProvider(metric_readers=[metric_reader])
metrics.set_meter_provider(meter_provider)
telemetry_module._TRACER_PROVIDER = tracer_provider
self.collector = InMemoryTelemetryCollector(span_exporter, metric_reader)
self._tracer_provider = tracer_provider
self._meter_provider = meter_provider
def shutdown(self) -> None:
telemetry_module._TRACER_PROVIDER = None
self._tracer_provider.shutdown()
self._meter_provider.shutdown()

2
tests/integration/telemetry/conftest.py
View file

@ -15,6 +15,7 @@ from tests.integration.fixtures.common import instantiate_llama_stack_client
from tests.integration.telemetry.collectors import InMemoryTelemetryManager, OtlpHttpTestCollector
# TODO: Fix this to work with Automatic Instrumentation
@pytest.fixture(scope="session")
def telemetry_test_collector():
stack_mode = os.environ.get("LLAMA_STACK_TEST_STACK_CONFIG_TYPE", "library_client")
@ -48,6 +49,7 @@ def telemetry_test_collector():
manager.shutdown()
# TODO: Fix this to work with Automatic Instrumentation
@pytest.fixture(scope="session")
def llama_stack_client(telemetry_test_collector, request):
"""Ensure telemetry collector is ready before initializing the stack client."""