3 commits

Author	SHA1	Message	Date
Alexey Rybak	8f57b08f2c	fix(build): always pass path when no template/config provided (#1982) ... # What does this PR do? Fixes a crash that occurred when building a stack as a container image via the interactive wizard without supplying --template or --config. - Root cause: template_or_config was None; only the container path relies on that parameter, which later reaches subprocess.run() and triggers `TypeError: expected str, bytes or os.PathLike object, not NoneType.` - Change: in `_run_stack_build_command_from_build_config` we now fall back to the freshly‑written build‑spec file whenever both optional sources are missing. Also adds a spy‑based unit test that asserts a valid string path is passed to build_image() for container builds. ### Closes #1976 ## Test Plan - New unit test: test_build_path.py. Monkey‑patches build_image, captures the fourth argument, and verifies it is a real path - Manual smoke test: ``` llama stack build --image-type container # answer wizard prompts ``` Build proceeds into Docker without raising the previous TypeError. ## Future Work Harmonise `build_image` arguments so every image type receives the same inputs, eliminating this asymmetric special‑case.	2025-04-17 10:20:43 +02:00
Sébastien Han	389767010b	feat: ability to execute external providers (#1672) ... # What does this PR do? Providers that live outside of the llama-stack codebase are now supported. A new property `external_providers_dir` has been added to the main config and can be configured as follow: ``` external_providers_dir: /etc/llama-stack/providers.d/ ``` Where the expected structure is: ``` providers.d/ inference/ custom_ollama.yaml vllm.yaml vector_io/ qdrant.yaml ``` Where `custom_ollama.yaml` is: ``` adapter: adapter_type: custom_ollama pip_packages: ["ollama", "aiohttp"] config_class: llama_stack_ollama_provider.config.OllamaImplConfig module: llama_stack_ollama_provider api_dependencies: [] optional_api_dependencies: [] ``` Obviously the package must be installed on the system, here is the `llama_stack_ollama_provider` example: ``` $ uv pip show llama-stack-ollama-provider Using Python 3.10.16 environment at: /Users/leseb/Documents/AI/llama-stack/.venv Name: llama-stack-ollama-provider Version: 0.1.0 Location: /Users/leseb/Documents/AI/llama-stack/.venv/lib/python3.10/site-packages Editable project location: /private/var/folders/mq/rnm5w_7s2d3fxmtkx02knvhm0000gn/T/tmp.ZBHU5Ezxg4/ollama/llama-stack-ollama-provider Requires: Required-by: ``` Closes: https://github.com/meta-llama/llama-stack/issues/658 Signed-off-by: Sébastien Han <seb@redhat.com>	2025-04-09 10:30:41 +02:00
Sébastien Han	7d9adf22ad	refactor: move missing tests to test directory (#1892) ... Move the test_context.py under the main tests directory, and fix the code. The problem was that the function captures the initial values of the context variables and then restores those same initial values before each iteration. This means that any modifications made to the context variables during iteration are lost when the next iteration starts. Error was: ``` ====================================================== FAILURES ======================================================= ______________________________________ test_preserve_contexts_across_event_loops ______________________________________ @pytest.mark.asyncio async def test_preserve_contexts_across_event_loops(): """ Test that context variables are preserved across event loop boundaries with nested generators. This simulates the real-world scenario where: 1. A new event loop is created for each streaming request 2. The async generator runs inside that loop 3. There are multiple levels of nested generators 4. Context needs to be preserved across these boundaries """ # Create context variables request_id = ContextVar("request_id", default=None) user_id = ContextVar("user_id", default=None) # Set initial values # Results container to verify values across thread boundaries results = [] # Inner-most generator (level 2) async def inner_generator(): # Should have the context from the outer scope yield (1, request_id.get(), user_id.get()) # Modify one context variable user_id.set("user-modified") # Should reflect the modification yield (2, request_id.get(), user_id.get()) # Middle generator (level 1) async def middle_generator(): inner_gen = inner_generator() # Forward the first yield from inner item = await inner_gen.__anext__() yield item # Forward the second yield from inner item = await inner_gen.__anext__() yield item request_id.set("req-modified") # Add our own yield with both modified variables yield (3, request_id.get(), user_id.get()) # Function to run in a separate thread with a new event loop def run_in_new_loop(): # Create a new event loop for this thread loop = asyncio.new_event_loop() asyncio.set_event_loop(loop) try: # Outer generator (runs in the new loop) async def outer_generator(): request_id.set("req-12345") user_id.set("user-6789") # Wrap the middle generator wrapped_gen = preserve_contexts_async_generator(middle_generator(), [request_id, user_id]) # Process all items from the middle generator async for item in wrapped_gen: # Store results for verification results.append(item) # Run the outer generator in the new loop loop.run_until_complete(outer_generator()) finally: loop.close() # Run the generator chain in a separate thread with a new event loop with ThreadPoolExecutor(max_workers=1) as executor: future = executor.submit(run_in_new_loop) future.result() # Wait for completion # Verify the results assert len(results) == 3 # First yield should have original values assert results[0] == (1, "req-12345", "user-6789") # Second yield should have modified user_id assert results[1] == (2, "req-12345", "user-modified") # Third yield should have both modified values > assert results[2] == (3, "req-modified", "user-modified") E AssertionError: assert (3, 'req-modified', 'user-6789') == (3, 'req-modified', 'user-modified') E E At index 2 diff: 'user-6789' != 'user-modified' E E Full diff: E ( E 3, E 'req-modified', E - 'user-modified', E + 'user-6789', E ) tests/unit/distribution/test_context.py:155: AssertionError -------------------------------------------------- Captured log call -------------------------------------------------- ERROR asyncio:base_events.py:1758 Task was destroyed but it is pending! task: <Task pending name='Task-7' coro=<<async_generator_athrow without __name__>()>> ================================================== warnings summary =================================================== .venv/lib/python3.10/site-packages/pydantic/fields.py:1042 /Users/leseb/Documents/AI/llama-stack/.venv/lib/python3.10/site-packages/pydantic/fields.py:1042: PydanticDeprecatedSince20: Using extra keyword arguments on `Field` is deprecated and will be removed. Use `json_schema_extra` instead. (Extra keys: 'contentEncoding'). Deprecated in Pydantic V2.0 to be removed in V3.0. See Pydantic V2 Migration Guide at https://errors.pydantic.dev/2.10/migration/ warn( -- Docs: https://docs.pytest.org/en/stable/how-to/capture-warnings.html =============================================== short test summary info =============================================== FAILED tests/unit/distribution/test_context.py::test_preserve_contexts_across_event_loops - AssertionError: assert (3, 'req-modified', 'user-6789') == (3, 'req-modified', 'user-modified') At index 2 diff: 'user-6789' != 'user-modified' Full diff: ( 3, 'req-modified', - 'user-modified', + 'user-6789', ) ``` [//]: # (## Documentation) Signed-off-by: Sébastien Han <seb@redhat.com>	2025-04-08 18:54:00 -07:00