You need to enable JavaScript to run this app.
优惠活动
大模型
产品
解决方案
定价
更多

Flask后端Python Asyncio后台定时任务启停报错排查与修复

avatar
API 探路者
2026-7-8
Flask后台定时任务启动/停止触发500错误的修复方案

问题场景

开发Python Flask后端应用，需实现后台定时任务（如每分钟自动更新数据），但调用启动/停止接口时出现以下问题：

  • 调用AsyncPeriodictUtils.start_task()：后端返回500错误，但定时任务实际已启动
  • 调用AsyncPeriodictUtils.stop_task()：后端返回500错误，且定时任务无法停止

相关代码

view_func.py 

from flask import Blueprint, request

from async_periodic_job import AsyncPeriodictUtils

infras_bp = Blueprint("infras", __name__)

@infras_bp.route("/infras/autosync", methods=["PUT"])
def auto_sync():
    args = request.args
    check_required_params(args, ["autosnyc"])
    autosnyc = args.get("autosnyc", default="false").lower() == "true"
    if autosnyc:
        AsyncPeriodictUtils.start_task()
    else:
        AsyncPeriodictUtils.stop_task()
    return "success", 200

async_periodic_job.py 

import asyncio
from typing import Callable

from utils import logger

SECOND = 1
MINUTE = 60 * SECOND

logger.debug(f"Import {__file__}")

periodic_jobs = dict()
task_instance = None

# Get default event loop in main thread
loop = asyncio.get_event_loop()


class AsyncPeriodictUtils:

    @staticmethod
    async def run_jobs() -> None:
        while True:
            await asyncio.sleep(10)
            logger.info(f"Called run_jobs periodicly.")
            logger.info(f"periodic_jobs: {periodic_jobs.keys()}")
            for func_name, function in periodic_jobs.items():
                function()
                logger.info(f"Called function '{func_name}' periodicly.")

    @classmethod
    def create_task(cls) -> None:
        global task_instance, loop
        task_instance = loop.create_task(cls.run_jobs())

    @staticmethod
    async def cancel_task() -> None:
        global task_instance
        if task_instance:
            task_instance.cancel()
            try:
                await task_instance
            except asyncio.CancelledError:
                logger.info("Async periodic task has been cancelled.")
            task_instance = None
        else:
            logger.warning("Async periodic task has not been started yet.")

    @classmethod
    def start_task(cls) -> None:
        cls.create_task()
        global loop
        try:
            loop.run_until_complete(task_instance)
        except asyncio.CancelledError:
            pass
        logger.info("Async Periodic jobs launched.")

    @classmethod
    def stop_task(cls) -> None:
        global loop
        try:
            loop.run_until_complete(cls.cancel_task())
        except asyncio.CancelledError:
            pass
        logger.info("Async Periodic jobs terminated.")

    @classmethod
    def add_job(cls, function: Callable) -> None:
        if function.__name__ in periodic_jobs:
            return
        periodic_jobs.update({function.__name__: function})
        logger.info(f"Added function {function.__name__} in periodic jobs.")
        global task_instance
        if not task_instance:
            logger.info(f"Auto enable periodic jobs.")
            cls.start_task()

    @classmethod
    def remove_job(cls, function: Callable) -> None:
        if function.__name__ not in periodic_jobs:
            logger.warning(
                f"function {function.__name__} not in periodic jobs.")
            return
        periodic_jobs.pop(function.__name__)
        logger.info(f"Removed function {function.__name__} in periodic jobs.")
        if not periodic_jobs:
            logger.info(f"Periodic jobs list clear, auto terminate.")
            cls.stop_task()

错误日志

启动任务时的500错误日志 

[2023-10-13-06:06:19][DAAT][ERROR][__init__.py] [Error] stack:
[2023-10-13-06:06:19][DAAT][ERROR][__init__.py] Traceback (most recent call last):
  File "/opt/conda/lib/python3.10/site-packages/flask/app.py", line 1820, in full_dispatch_request
    rv = self.dispatch_request()
  File "/opt/conda/lib/python3.10/site-packages/flask/app.py", line 1796, in dispatch_request
    return self.ensure_sync(self.view_functions[rule.endpoint])(**view_args)
  File "/opt/conda/lib/python3.10/site-packages/daat-1.0.0-py3.10.egg/daat/app/routes/infras.py", line 134, in auto_sync
    AsyncPeriodictUtils.start_task()
  File "/opt/conda/lib/python3.10/site-packages/daat-1.0.0-py3.10.egg/daat/infras/async_periodic_job.py", line 53, in start_task
    loop.run_until_complete(task_instance)
  File "/opt/conda/lib/python3.10/asyncio/base_events.py", line 625, in run_until_complete
    self._check_running()
  File "/opt/conda/lib/python3.10/asyncio/base_events.py", line 584, in _check_running
    raise RuntimeError('This event loop is already running')
RuntimeError: This event loop is already running

[2023-10-13-06:06:19][werkzeug][INFO][_internal.py] 172.23.0.3 - - [13/Oct/2023 06:06:19] "PUT /infras/autosync?autosnyc=true HTTP/1.1" 500

启动后任务运行日志 

[2023-10-13-06:06:23][DAAT][INFO][async_periodic_job.py] Called run_jobs periodicly.
[2023-10-13-06:06:23][DAAT][INFO][async_periodic_job.py] periodic_jobs: dict_keys([])
[2023-10-13-06:06:33][DAAT][INFO][async_periodic_job.py] Called run_jobs periodicly.
[2023-10-13-06:06:33][DAAT][INFO][async_periodic_job.py] periodic_jobs: dict_keys([])

停止任务时的500错误日志 

[2023-10-13-06:06:35][DAAT][ERROR][__init__.py] [Error] stack:
[2023-10-13-06:06:35][DAAT][ERROR][__init__.py] Traceback (most recent call last):
  File "/opt/conda/lib/python3.10/site-packages/flask/app.py", line 1820, in full_dispatch_request
    rv = self.dispatch_request()
  File "/opt/conda/lib/python3.10/site-packages/flask/app.py", line 1796, in dispatch_request
    return self.ensure_sync(self.view_functions[rule.endpoint])(**view_args)
  File "/opt/conda/lib/python3.10/site-packages/daat-1.0.0-py3.10.egg/daat/app/routes/infras.py", line 136, in auto_sync
    AsyncPeriodictUtils.stop_task()
  File "/opt/conda/lib/python3.10/site-packages/daat-1.0.0-py3.10.egg/daat/infras/async_periodic_job.py", line 62, in stop_task
    loop.run_until_complete(cls.cancel_task())
  File "/opt/conda/lib/python3.10/asyncio/base_events.py", line 625, in run_until_complete
    self._check_running()
  File "/opt/conda/lib/python3.10/asyncio/base_events.py", line 584, in _check_running
    raise RuntimeError('This event loop is already running')
RuntimeError: This event loop is already running

/opt/conda/lib/python3.10/site-packages/flask/app.py:1822: RuntimeWarning: coroutine 'AsyncPeriodictUtils.cancel_task' was never awaited
  rv = self.handle_user_exception(e)
RuntimeWarning: Enable tracemalloc to get the object allocation traceback
[2023-10-13-06:06:35][werkzeug][INFO][_internal.py] 172.23.0.3 - - [13/Oct/2023 06:06:35] "PUT /infras/autosync?autosnyc=false HTTP/1.1" 500 -

停止后任务仍运行的日志 

[2023-10-13-06:06:43][DAAT][INFO][async_periodic_job.py] Called run_jobs periodicly.
[2023-10-13-06:06:43][DAAT][INFO][async_periodic_job.py] periodic_jobs: dict_keys([])
[2023-10-13-06:06:53][DAAT][INFO][async_periodic_job.py] Called run_jobs periodicly.
[2023-10-13-06:06:53][DAAT][INFO][async_periodic_job.py] periodic_jobs: dict_keys([])

预期结果

  • 调用AsyncPeriodictUtils.start_task()：成功启动定时任务，后端返回200，无500错误
  • 调用AsyncPeriodictUtils.stop_task()：成功终止定时任务，后端返回200，无500错误

解决方案

问题根源

  1. Flask运行时，主线程的事件循环已经处于运行状态，调用loop.run_until_complete()会触发RuntimeError: This event loop is already running，因为该方法要求事件循环未启动，且会阻塞等待任务完成（但定时任务是无限循环，无法完成）。
  2. 停止任务时，直接调用loop.run_until_complete(cls.cancel_task())同样触发上述错误，且协程未被正确执行，导致任务无法停止。

修改后的async_periodic_job.py关键代码 

import asyncio
from typing import Callable

from utils import logger

SECOND = 1
MINUTE = 60 * SECOND

logger.debug(f"Import {__file__}")

periodic_jobs = dict()
task_instance = None

# Get default event loop in main thread
loop = asyncio.get_event_loop()


class AsyncPeriodictUtils:

    @staticmethod
    async def run_jobs() -> None:
        while True:
            await asyncio.sleep(10)
            logger.info(f"Called run_jobs periodicly.")
            logger.info(f"periodic_jobs: {periodic_jobs.keys()}")
            for func_name, function in periodic_jobs.items():
                function()
                logger.info(f"Called function '{func_name}' periodicly.")

    @classmethod
    def create_task(cls) -> None:
        global task_instance, loop
        # 防止重复创建任务
        if task_instance is not None and not task_instance.done():
            logger.warning("Async periodic task is already running.")
            return
        task_instance = loop.create_task(cls.run_jobs())

    @staticmethod
    async def cancel_task() -> None:
        global task_instance
        if task_instance:
            task_instance.cancel()
            try:
                await task_instance
            except asyncio.CancelledError:
                logger.info("Async periodic task has been cancelled.")
            task_instance = None
        else:
            logger.warning("Async periodic task has not been started yet.")

    @classmethod
    def start_task(cls) -> None:
        cls.create_task()
        logger.info("Async Periodic jobs launched.")

    @classmethod
    def stop_task(cls) -> None:
        global loop
        # 将cancel_task协程提交到已有事件循环执行，而非阻塞等待
        loop.create_task(cls.cancel_task())
        logger.info("Async Periodic jobs termination requested.")

    # 其余add_job、remove_job方法保持不变
    @classmethod
    def add_job(cls, function: Callable) -> None:
        if function.__name__ in periodic_jobs:
            return
        periodic_jobs.update({function.__name__: function})
        logger.info(f"Added function {function.__name__} in periodic jobs.")
        global task_instance
        if not task_instance or (task_instance and task_instance.done()):
            logger.info(f"Auto enable periodic jobs.")
            cls.start_task()

    @classmethod
    def remove_job(cls, function: Callable) -> None:
        if function.__name__ not in periodic_jobs:
            logger.warning(
                f"function {function.__name__} not in periodic jobs.")
            return
        periodic_jobs.pop(function.__name__)
        logger.info(f"Removed function {function.__name__} in periodic jobs.")
        if not periodic_jobs:
            logger.info(f"Periodic jobs list clear, auto terminate.")
            cls.stop_task()

修改说明

  1. start_task方法：移除loop.run_until_complete(task_instance)调用，因为loop.create_task()已经将定时任务加入到运行中的事件循环，无需阻塞等待任务完成。同时在create_task中增加重复启动检查，避免创建多个相同任务。
  2. stop_task方法：用loop.create_task(cls.cancel_task())替代loop.run_until_complete(cls.cancel_task())，将取消任务的协程提交到已有事件循环异步执行，不会阻塞请求线程，也不会触发事件循环已运行的错误。
  3. add_job方法：更新任务存在性判断，确保只有在任务未启动或已完成时才自动启动任务。

验证效果

修改后调用接口：

  • 启动任务：后端返回200，定时任务正常运行，无500错误
  • 停止任务：后端返回200，定时任务停止运行，无500错误

内容的提问来源于stack exchange，提问作者stevezkw

相关产品推荐
方舟 Agent Plan

超全模态模型 × Harness 升级，最新支持 Deepseek-V4.1-Flash、GLM-5.3 系列、Doubao-Seedream-5.0-pro、Kimi-K3 (部分)， 限时 9.9 元起

产品介绍相关文档
相关产品推荐
方舟 Agent Plan

超全模态模型 × Harness 升级，最新支持 Deepseek-V4.1-Flash、GLM-5.3 系列、Doubao-Seedream-5.0-pro、Kimi-K3 (部分)， 限时 9.9 元起

产品介绍相关文档
火山引擎 最新活动
全面开放
Doubao-Seedance-2.5全新模型升级

模型再升级,30秒超长叙事, 模态参考扩容

HOT
方舟Codingplan

模型自由，工具不限，最新支持 Deepseek-V4 系列、GLM-5.3 系列

HOT
方舟Agent plan

超全模态模型 × Harness 升级，最新支持 Deepseek-V4.1-Flash、GLM-5.3 系列

HOT
新用户特惠专场

大模型19元起，Al应用9.9元畅享，新人首购爆款尽享优惠

最近更新时间：2026.07.08 20:24:50