Files
BanGUI/backend/app/tasks/geo_re_resolve.py
Lukas 52f237d5d4 Make background tasks idempotent - prevent duplicate bans on retry
CRITICAL FIX: Background tasks (especially blocklist_import) crashed mid-execution,
leaving partial state. On retry, the same bans were applied again, causing duplicates.

Solution: Content-hash based operation tracking for blocklist imports:
- Added import_runs table (migration 6) to track operations by source + content hash
- Before banning, check if this exact content has already been imported
- If completed: skip banning (already done), optionally re-warm cache
- If new or failed: proceed with ban and mark as completed or failed

Changes:
- Database: Migration 6 adds import_runs table with operation state tracking
- Model: Added ImportRunEntry for import run records
- Repository: New import_run_repo module with CRUD operations
- Workflow: Updated blocklist_import_workflow to check operation history before banning
- Dependencies: Registered import_run_repo for dependency injection
- Tests: Added test_import_source_idempotent_on_retry and test_import_source_different_content_not_reused
- Documentation: Added Task Idempotency section to Backend-Development.md

Verification:
- All 7 import tests pass (5 existing + 2 new idempotency tests)
- Type checking: mypy --strict 
- Linting: ruff 
- No API changes, backwards compatible via automatic migration

Fixes: Background tasks not idempotent #CRITICAL

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
2026-04-30 21:54:14 +02:00

119 lines
4.2 KiB
Python

"""Geo re-resolve background task.
Registers an APScheduler job that periodically retries IP addresses in the
``geo_cache`` table whose ``country_code`` is ``NULL``. These are IPs that
previously failed to resolve (e.g. due to ip-api.com rate limiting) and were
recorded as negative entries.
The task runs every 10 minutes. On each invocation it:
1. Queries all ``NULL``-country rows from ``geo_cache``.
2. Clears the in-memory negative cache so those IPs are eligible for a fresh
API attempt.
3. Delegates to :meth:`~app.services.geo_cache.GeoCache.lookup_batch` which
already handles rate-limit throttling and retries.
4. Logs how many IPs were retried and how many resolved successfully.
"""
from __future__ import annotations
from typing import TYPE_CHECKING
import structlog
from app.tasks.db import task_db
from app.tasks.timeout_utils import run_with_timeout
from app.utils.runtime_state import get_effective_settings
if TYPE_CHECKING:
from aiohttp import ClientSession
from fastapi import FastAPI
from app.config import Settings
from app.services.geo_cache import GeoCache
log: structlog.stdlib.BoundLogger = structlog.get_logger()
#: How often the re-resolve job fires (seconds). 10 minutes.
GEO_RE_RESOLVE_INTERVAL: int = 600
#: Stable APScheduler job ID — ensures re-registration replaces, not duplicates.
JOB_ID: str = "geo_re_resolve"
#: Maximum seconds to allow for geo re-resolve to complete.
TASK_TIMEOUT_SECONDS: int = 120
async def _run_re_resolve_with_resources(
geo_cache: GeoCache, settings: Settings, http_session: ClientSession
) -> None:
"""Query NULL-country IPs from the database and re-resolve them.
Args:
geo_cache: The application's GeoCache instance.
settings: The resolved application settings used for database access.
http_session: The shared aiohttp session used for external lookups.
"""
async def _do_re_resolve() -> None:
async with task_db(settings) as db:
# Fetch all IPs with NULL country_code from the persistent cache.
unresolved_ips = await geo_cache.get_unresolved_ips(db)
if not unresolved_ips:
log.debug("geo_re_resolve_skip", reason="no_unresolved_ips")
return
log.info("geo_re_resolve_start", unresolved=len(unresolved_ips))
# Clear the negative cache so these IPs are eligible for fresh API calls.
await geo_cache.clear_neg_cache()
# lookup_batch handles throttling, retries, and persistence when db is
# passed. This is a background task so DB writes are allowed.
results = await geo_cache.lookup_batch(unresolved_ips, http_session, db=db)
resolved_count: int = sum(
1 for info in results.values() if info.country_code is not None
)
log.info(
"geo_re_resolve_complete",
retried=len(unresolved_ips),
resolved=resolved_count,
)
await run_with_timeout("geo_re_resolve", _do_re_resolve(), TASK_TIMEOUT_SECONDS)
async def _run_re_resolve(app: FastAPI) -> None:
geo_cache: GeoCache = app.state.geo_cache
await _run_re_resolve_with_resources(
geo_cache, get_effective_settings(app), app.state.http_session
)
def register(app: FastAPI) -> None:
"""Add (or replace) the geo re-resolve job in the application scheduler.
Must be called after the scheduler has been started (i.e., inside the
lifespan handler, after ``scheduler.start()``).
The first invocation is deferred by one full interval so the initial
blocklist prewarm has time to finish before re-resolve kicks in.
Args:
app: The :class:`fastapi.FastAPI` application instance whose
``app.state.scheduler`` will receive the job.
"""
geo_cache: GeoCache = app.state.geo_cache
settings = get_effective_settings(app)
app.state.scheduler.add_job(
_run_re_resolve_with_resources,
trigger="interval",
seconds=GEO_RE_RESOLVE_INTERVAL,
kwargs={"geo_cache": geo_cache, "settings": settings, "http_session": app.state.http_session},
id=JOB_ID,
replace_existing=True,
)
log.info("geo_re_resolve_scheduled", interval_seconds=GEO_RE_RESOLVE_INTERVAL)