chore(misc): make tests and starter faster

tests/integration/conftest.py

@@ -9,12 +9,15 @@ import os
 import platform
 import textwrap
 import time
+import warnings
 
 import pytest
 from dotenv import load_dotenv
 
 from llama_stack.log import get_logger
 
+from .fixtures.common import instantiate_llama_stack_client
+
 logger = get_logger(__name__, category="tests")
 
 
@@ -27,6 +30,20 @@ def pytest_runtest_makereport(item, call):
         item.was_xfail = getattr(report, "wasxfail", False)
 
 
+def pytest_sessionstart(session):
+    # stop macOS from complaining about duplicate OpenMP libraries
+    os.environ["KMP_DUPLICATE_LIB_OK"] = "TRUE"
+
+    # pull client instantiation to session start so all the complex logs during initialization
+    # don't clobber the test one-liner outputs
+    print("instantiating llama_stack_client")
+    start_time = time.time()
+    with warnings.catch_warnings():
+        warnings.filterwarnings("ignore", category=DeprecationWarning)
+        session._llama_stack_client = instantiate_llama_stack_client(session)
+    print(f"llama_stack_client instantiated in {time.time() - start_time:.3f}s")
+
+
 def pytest_runtest_teardown(item):
     # Check if the test actually ran and passed or failed, but was not skipped or an expected failure (xfail)
     outcome = getattr(item, "execution_outcome", None)