Flask Academy · 课时

跟踪结果并处理失败

轮询任务状态并在出错时重试

第 4 / 4 课13 个步骤

跟踪结果并处理失败 是 CoddyKit 上的免费 Flask Academy 课时。这是第 4 节课，共 4 节。你可以在下方免费阅读本课时的完整内容 — 然后在浏览器中使用内置代码编辑器和全天候 AI 导师进行实践。这是 Flask Academy 学习路径的一部分，你的进度在网页和 CoddyKit 应用中同步。 Flask Academy 课程共包含 4 节课。

本课时的部分内容尚未翻译，以英文显示。

The AsyncResult Handle

When you enqueue a task you get an AsyncResult. Save its id so you can check on the job from any later request.

job = add.delay(2, 3)
ticket = job.id

Check the State

Each job has a state like PENDING, STARTED, SUCCESS, or FAILURE. You read it to know whether the work is done.

res = add.AsyncResult(ticket)
print(res.state)

Read the Result

Once a job succeeds, its return value is stored in the backend. Access it through result on the AsyncResult.

if res.ready():
    print(res.result)

Poll, Do Not Block

In a web view, never call get() and wait. Instead return the task id and let the client poll a status endpoint.

A Status Endpoint

Expose a route that takes a task id and returns its state and result. The frontend hits it every few seconds until done.

@app.get("/status/<tid>")
def status(tid):
    r = add.AsyncResult(tid)
    return {"state": r.state}

Failures Happen

External calls time out and code raises errors. When a task throws, Celery marks it FAILURE and records the exception.

Automatic Retries

Make a task retry itself on transient errors. Set max_retries and a delay so flaky calls get a few more chances.

@celery.task(bind=True, max_retries=3)
def fetch(self):
    ...

Back Off Between Tries

Retrying instantly can hammer a struggling service. Use a growing delay, called backoff, so each retry waits longer.

Trigger a Retry

Inside the task, catch the error and call self.retry. Celery re-queues the job and counts it against max_retries.

try:
    do_work()
except Exception as e:
    self.retry(exc=e, countdown=5)

Set Time Limits

A stuck task should not run forever. A time limit kills tasks that overrun, freeing the worker for other jobs.

Result Expiry

Stored results take space. Celery expires old results after a while, so design clients to fetch them promptly. ⏱️

Quick Check

Your task calls a flaky external API.

在交互式课程中回答

Recap

Track jobs by id, poll a status endpoint for state and result, and handle failures with retries, backoff, and time limits. ✅

免费开始

用 AI 导师学习 Python — 免费

在浏览器中编写并运行真实代码，获得全天候 AI 导师的即时帮助，并在网页或应用中继续学习。

课程: 30
课程: 120

免费开始学习查看完整课程

常见问题解答

「跟踪结果并处理失败」课时是免费的吗？

是的 — 「跟踪结果并处理失败」的完整文本可在网页上免费阅读。要进行交互式练习（内置代码编辑器和全天候 AI 导师）并解锁 Flask Academy 课程的其余内容，请升级到 CoddyKit PRO。 Flask Academy 课程共包含 4 节课。

「跟踪结果并处理失败」这节课中我会学到什么？

轮询任务状态并在出错时重试你通过在浏览器中直接运行的动手代码来练习 Flask Academy，全天候 AI 导师会在你学习这节课的过程中回答你的问题。

学习 Flask Academy 需要有经验吗？

无需任何先前经验。CoddyKit 上的 Flask Academy 课程适合初学者到高级学习者，你可以从这里开始或从头开始，按照自己的节奏学习。这是第 4 节课，共 4 节。

「跟踪结果并处理失败」课时需要多长时间？

大多数 CoddyKit 课程大约需要 5–10 分钟。每节课都很精短且互动，所以你能稳步进步，并在网页和应用中从离开的地方继续。

我能在这节 Flask Academy 课中编写并运行代码吗？

能。每节 Flask Academy 课都包含内置代码编辑器，你可以在浏览器中直接编写并运行真实代码，并获得即时 AI 反馈 — 无需本地设置。

此课程中的所有课时

← 返回 Flask Academy