mirror of
https://github.com/bytedance/deer-flow.git
synced 2026-08-01 19:06:01 +00:00
7852421c68
Adds the three remaining adapters plus the poller. Nothing is wired yet -- the composition root is the next commit -- so this is additive and the legacy `app/scheduler/service.py` still serves production. `run_launcher.py` is the pivot of the whole slice. The Gateway signals a busy thread two ways -- `ConflictError` from the run manager, or an `HTTPException(409)` from the route-level path -- which is why the legacy scheduler service imported fastapi to tell them apart. Both are one domain fact, and saying so here is what lets that import disappear without the busy/failed distinction disappearing with it. Everything else becomes `LaunchFailedError`, because the port promises the domain that nothing but its two errors escapes. `CancelledError` is deliberately not caught: shutdown is control flow, not a launch outcome. `thread_lookup.py` narrows `ThreadMetaStore` to the one question this context asks. `require_existing=True` is load-bearing -- the store's default treats an absent row as accessible, which is right for a thread not yet written and wrong for binding a task to it. Both inherit their port explicitly, matching every other adapter in the codebase including feedback's own anti-corruption layer, and both carry the TODO naming the published contract that would replace them once the upstream context has been through a slice of its own. `run_outcome_mapping.py` implements no port: it is the inbound translation the composition root will install on the completion hook, and it owns the filtering the legacy hook did inline. Returning None means "this run is none of the schedule context's business", so the service is simply never called and needs no guard clauses. `poller.py` keeps the two behaviours the legacy loop got right: a failing poll must not end the loop (one transient "database is locked" used to stop scheduling for the rest of the process life), and reconciliation must not block startup. One deliberate behaviour change: the legacy `start()` swept stale runs and stuck once-tasks under separate try/excepts, so the first failing did not stop the second. `reconcile_on_startup` is one call that lets failures propagate -- the domain's position is that fatality is the caller's policy -- so the poller's single except means a failed first sweep now skips the second. Both end up logged and non-fatal, as before. Tests: 50 new cases across the four modules, each port method called and asserted on its return value. That is not decoration: inheriting a Protocol means a misspelled method silently inherits its `...` body and returns None, so the suite was verified by mutation -- renaming `launch` and `exists_for_user` turns 16 and 6 cases red respectively.
85 lines
3.4 KiB
Python
85 lines
3.4 KiB
Python
"""Driving adapter -- the clock that asks the schedule service to work.
|
|
|
|
This is the only part of the scheduler that knows about time passing. It owns
|
|
*when* `ScheduleService.run_once` is called and what happens when a poll fails;
|
|
it owns nothing about what a poll means. Everything the old
|
|
`app/scheduler/service.py` mixed into its loop -- overlap policy, lease
|
|
semantics, budget accounting -- now lives in the domain and reaches this file
|
|
only as one awaited call.
|
|
|
|
Two behaviours here are load-bearing:
|
|
|
|
- **A failing poll must not end the loop.** A transient error (SQLite's
|
|
"database is locked" is the realistic one) would otherwise stop every
|
|
scheduled task for the rest of the process life, silently.
|
|
- **Startup reconciliation must not block startup.** The service lets
|
|
reconcile failures propagate on purpose -- whether they are fatal is the
|
|
caller's policy -- and this caller's policy is to log and keep scheduling.
|
|
A gateway refusing to start over leftover rows is worse than one running
|
|
with them.
|
|
"""
|
|
|
|
from __future__ import annotations
|
|
|
|
import asyncio
|
|
import logging
|
|
from datetime import UTC, datetime
|
|
from typing import TYPE_CHECKING
|
|
|
|
if TYPE_CHECKING:
|
|
from deerflow.domain.schedule.service import ScheduleService
|
|
|
|
logger = logging.getLogger(__name__)
|
|
|
|
RESTART_ERROR = "interrupted: gateway restarted before the run reached a terminal state"
|
|
|
|
|
|
class SchedulePoller:
|
|
"""Runs `ScheduleService.run_once` on an interval until stopped."""
|
|
|
|
def __init__(self, service: ScheduleService, *, poll_interval_seconds: float) -> None:
|
|
self._service = service
|
|
self._poll_interval_seconds = poll_interval_seconds
|
|
self._task: asyncio.Task | None = None
|
|
self._stop = asyncio.Event()
|
|
|
|
async def start(self) -> None:
|
|
"""Reconcile what a crash left behind, then begin polling.
|
|
|
|
Idempotent: a second call while running is a no-op, so a caller cannot
|
|
end up with two loops claiming the same tasks.
|
|
"""
|
|
if self._task is not None:
|
|
return
|
|
try:
|
|
stale_runs, stuck_tasks = await self._service.reconcile_on_startup(error=RESTART_ERROR)
|
|
if stale_runs:
|
|
logger.warning("Marked %d stale scheduled task run(s) as interrupted after restart", stale_runs)
|
|
if stuck_tasks:
|
|
logger.warning("Cancelled %d stuck once task(s) after restart", stuck_tasks)
|
|
except Exception:
|
|
logger.exception("Failed to reconcile scheduled tasks at startup; scheduling anyway")
|
|
self._stop.clear()
|
|
self._task = asyncio.create_task(self._run_loop())
|
|
|
|
async def stop(self) -> None:
|
|
"""Signal the loop and wait for the in-flight poll to finish."""
|
|
if self._task is None:
|
|
return
|
|
self._stop.set()
|
|
await self._task
|
|
self._task = None
|
|
|
|
async def _run_loop(self) -> None:
|
|
while not self._stop.is_set():
|
|
try:
|
|
await self._service.run_once(now=datetime.now(UTC))
|
|
except Exception:
|
|
logger.exception("Scheduled task poll failed; retrying next interval")
|
|
try:
|
|
# Waiting on the stop event rather than sleeping keeps shutdown
|
|
# prompt at a production-sized interval.
|
|
await asyncio.wait_for(self._stop.wait(), timeout=self._poll_interval_seconds)
|
|
except TimeoutError:
|
|
continue
|