Files
neuronetz-gateway/src/neuronetz_gateway/lifespan.py
Stephan Berbig 653e03bf29
Some checks failed
CI / ruff (push) Has been cancelled
CI / mypy --strict (push) Has been cancelled
CI / pytest (push) Has been cancelled
CI / bandit (push) Has been cancelled
CI / pip-audit (push) Has been cancelled
proxy: multi-backend Ollama aggregation with per-model routing + failover
The gateway can now aggregate models across SEVERAL Ollama backends and
route each request to the correct one. Opt-in via OLLAMA_BACKENDS in .env
— single-backend deployments are unaffected (effective_backends()
synthesizes a single "default" backend from the legacy OLLAMA_BASE_URL /
OLLAMA_AUTH_TOKEN fields when the list is empty).

Behavior:
- Discovery polls EVERY configured backend in parallel each tick; the
  cache stores per-backend model lists plus a model → backends priority
  list (config order = priority order).
- /api/tags and /v1/models surface the DEDUPLICATED UNION of all
  backends' models.
- A request's model is looked up in the priority list and proxied to the
  FIRST backend that hosts it. If that backend errors on the request, the
  pipeline transparently fails over to the next backend that has the
  same model (the streaming-failover probes the first chunk before
  releasing the response, so we never serve partial bytes from a dead
  backend).
- No existence disclosure: a model not hosted by any backend yields the
  same generic 403 as "model not allowed" (SPEC §13.6 preserved).

Components:
- config.py: new BackendSpec model + ollama_backends list field + an
  effective_backends() helper.
- proxy/router.py (new): BackendRouter (clients_for_with_failover),
  build_http_clients() builds one httpx client per backend with its own
  auth headers, build_backend_headers() exposes the per-backend header
  composition for the CLI probe.
- proxy/discovery.py: DiscoveryCache.set_per_backend() + backends_for(),
  refresh_all_backends() polls all in parallel, discovery_loop_multi()
  replaces the single-backend loop in production; the legacy single-
  backend functions are kept for the dependency-override tests.
- proxy/pipeline.py: Pipeline accepts an optional router; the four proxy
  methods now retry against each candidate backend in priority order on
  transport error.
- lifespan.py: constructs the per-backend client dict, stores the router
  on app.state, launches discovery_loop_multi.
- deps.py: get_backend_router provider + BackendRouterDep type alias;
  get_pipeline passes the router into Pipeline.
- cli/manage.py: probe-ollama iterates every backend and reports per-
  backend status; list-models groups its output by backend and prints
  the union count + Redis cache size for sanity.
- .env.example + docker-compose.yml: document and pass through
  OLLAMA_BACKENDS with a real example.

Verified: ruff check (clean), mypy --strict src/ + tests/ (clean,
66 source files), pytest (60 passed + 39 skipped — same baseline as
before this change; integration tests are Docker-socket-gated).
2026-05-27 22:30:26 +02:00

181 lines
6.7 KiB
Python

"""Application lifespan: connect/dispose backends and run background tasks.
Startup connects Postgres + Redis + the upstream httpx client, builds the
argon2 hasher and the buffered audit writer, and launches the background tasks:
the model-discovery poller (SPEC §4.6) and the Postgres revocation NOTIFY
listener (SPEC §4.5). Connection failures are tolerated so ``/healthz`` always
serves; ``/readyz`` reports true readiness. All handles live on ``app.state``.
"""
from __future__ import annotations
import asyncio
import contextlib
from collections.abc import AsyncIterator
from contextlib import asynccontextmanager
from typing import TYPE_CHECKING
import httpx
import redis.asyncio as redis
from neuronetz_gateway.audit.writer import AuditWriter
from neuronetz_gateway.auth.hashing import build_hasher
from neuronetz_gateway.config import Settings, get_settings
from neuronetz_gateway.db.session import create_engine, create_session_factory
from neuronetz_gateway.observability.logging import get_logger
from neuronetz_gateway.proxy.discovery import DiscoveryCache, discovery_loop_multi
from neuronetz_gateway.proxy.ollama import OllamaClient
from neuronetz_gateway.proxy.router import BackendRouter, build_http_clients
from neuronetz_gateway.revocation import revocation_listener
if TYPE_CHECKING:
from fastapi import FastAPI
_log = get_logger("lifespan")
def _build_upstream_headers(settings: Settings) -> dict[str, str]:
"""Compose default headers for the upstream Ollama client.
If ``OLLAMA_AUTH_TOKEN`` is set, attach the configured auth header. The
scheme prefix (``Bearer``) is included only when the header is the standard
``Authorization``; for custom headers like ``X-API-Key`` the raw token is
sent. The SecretStr is unwrapped only here, never logged.
"""
headers: dict[str, str] = {"User-Agent": "neuronetz-gateway"}
if settings.ollama_auth_token is not None:
raw = settings.ollama_auth_token.get_secret_value().strip()
if raw:
header = settings.ollama_auth_header
if header.lower() == "authorization":
headers[header] = f"{settings.ollama_auth_scheme} {raw}".strip()
else:
headers[header] = raw
return headers
def _build_http_client(settings: Settings) -> httpx.AsyncClient:
"""Construct the shared httpx client used to reach Ollama."""
timeout = httpx.Timeout(
connect=settings.ollama_connect_timeout_s,
read=settings.ollama_read_timeout_s,
write=settings.ollama_read_timeout_s,
pool=settings.ollama_connect_timeout_s,
)
limits = httpx.Limits(max_connections=settings.ollama_max_connections)
return httpx.AsyncClient(
base_url=settings.ollama_base_url,
timeout=timeout,
limits=limits,
headers=_build_upstream_headers(settings),
)
@asynccontextmanager
async def lifespan(app: FastAPI) -> AsyncIterator[None]:
"""Manage startup/shutdown of all backends and background tasks."""
settings: Settings = get_settings()
app.state.settings = settings
app.state.hasher = build_hasher(settings)
app.state.discovery_cache = DiscoveryCache()
tasks: list[asyncio.Task[None]] = []
try:
engine = create_engine(settings)
app.state.db_engine = engine
app.state.db_sessionmaker = create_session_factory(engine)
except Exception as exc: # noqa: BLE001 - tolerate so /healthz still serves
_log.error("db_engine_init_failed", error=str(exc))
app.state.db_engine = None
app.state.db_sessionmaker = None
try:
app.state.redis = redis.from_url(settings.redis_url, decode_responses=True)
except Exception as exc: # noqa: BLE001 - tolerate so /healthz still serves
_log.error("redis_init_failed", error=str(exc))
app.state.redis = None
# Build per-backend upstream clients (one per entry in OLLAMA_BACKENDS, or
# a single "default" backend synthesized from the legacy OLLAMA_BASE_URL).
backend_clients, backend_order = build_http_clients(settings)
app.state.backend_http_clients = backend_clients
app.state.backend_order = backend_order
# ``http_client`` retains its single-client meaning for code paths (and
# tests) that haven't been migrated to the router yet: it is the FIRST
# backend's httpx client. New code should reach upstream via the router.
app.state.http_client = backend_clients[backend_order[0]]
app.state.backend_router = BackendRouter(
clients={name: OllamaClient(client) for name, client in backend_clients.items()},
order=backend_order,
discovery=app.state.discovery_cache,
)
_log.info("backends_configured", backends=backend_order)
audit_writer = AuditWriter(settings.audit_buffer_size, app.state.db_sessionmaker)
audit_writer.start()
app.state.audit_writer = audit_writer
# Background tasks (cancelled on shutdown).
tasks.append(
asyncio.create_task(
discovery_loop_multi(
backend_clients,
backend_order,
app.state.redis,
app.state.discovery_cache,
settings,
)
)
)
if app.state.redis is not None and app.state.db_sessionmaker is not None:
tasks.append(
asyncio.create_task(
revocation_listener(settings, app.state.redis, app.state.db_sessionmaker)
)
)
app.state.background_tasks = tasks
_log.info("gateway_startup_complete")
try:
yield
finally:
await _shutdown(app, tasks, audit_writer)
async def _shutdown(
app: FastAPI, tasks: list[asyncio.Task[None]], audit_writer: AuditWriter
) -> None:
"""Cancel background tasks and dispose of all backend handles."""
for task in tasks:
task.cancel()
for task in tasks:
with contextlib.suppress(asyncio.CancelledError):
await task
with contextlib.suppress(Exception):
await audit_writer.stop()
# Close every per-backend httpx client (the legacy `http_client` attr is
# one of these, so we only need to iterate the dict).
backend_clients: dict[str, httpx.AsyncClient] = getattr(
app.state, "backend_http_clients", {}
)
for client in backend_clients.values():
with contextlib.suppress(Exception):
await client.aclose()
redis_client = getattr(app.state, "redis", None)
if redis_client is not None:
with contextlib.suppress(Exception):
await redis_client.aclose()
engine = getattr(app.state, "db_engine", None)
if engine is not None:
with contextlib.suppress(Exception):
await engine.dispose()
_log.info("gateway_shutdown_complete")
__all__ = ["lifespan"]