Question

我刚刚开始使用Celery并对link_error处理有疑问。

这是我启动测试的方式：

res = tasks.p0.apply_async（（1,2,3），link = tasks.p1.s（（'AA'）），link_error = tasks.p0_fail.s（id ='qwerty'））< / p>

并且，为了调试p0_fail，我在p0中生成了不同的异常。这是p0_fail的定义：

@app.task
def p0_fail(wtf, *args, **kwargs):
    print("What the Failed's type is {}".format(type(wtf)))
    result = app.AsyncResult(wtf)
    print("Result type is {}".format(type(result)))
    code = kwargs['id']
    # exc = result.get(propagate=False, disable_sync_subtasks=False)
    print("Task {0} (code '{4}') raised exception: {1!r}\n{2!r} and is in '{3}' state".format(
        wtf.id, result.info, result.traceback, result.state, code))
    print("result.id is {} (type {})".format(result.id, type(result.id)))

产生以下输出：

[2018-06-07 10:48:40,943: WARNING/ForkPoolWorker-2] p0 got exception: division by zero
[2018-06-07 10:48:40,965: WARNING/ForkPoolWorker-2] What the Failed's type is <class 'celery.app.task.Context'>
[2018-06-07 10:48:40,965: WARNING/ForkPoolWorker-2] Result type is <class 'celery.result.AsyncResult'>
[2018-06-07 10:48:40,970: WARNING/ForkPoolWorker-2] Task 03b009af-dc35-46c5-8216-a6868e66c9be (code 'qwerty') raised exception: None None and is in 'PENDING' state
[2018-06-07 10:48:40,970: WARNING/ForkPoolWorker-2] result.id is <Context: {'lang': 'py', 'task': 'tasks.p0', 'id': '03b009af-dc35-46c5-8216-a6868e66c9be', [ ... ]}> (type <class 'celery.app.task.Context'>)

问题是：

无论p0_fail调用Exception，结果的（即celery.result.AsyncResult）属性.info / .result和.traceback都是'None'，而文档（http://docs.celeryproject.org/en/latest/reference/celery.result.html）表示应该有相应的异常实例和回溯。是bug还是功能？更新：作为第二和第三个参数传递的exc和traceback。
根据文档（http://docs.celeryproject.org/en/master/userguide/calling.html#linking-callbacks-errbacks）进行操作时，由于任务处于PENDING状态，我会在'result.get'调用中遇到死锁
- 如果任务因异常而失败，状态为PENDING？
- Celery的任务是否真的停止了，或者是否处于活跃/待定状态吃Celery的资源？
根据文档，result.id应该是任务的UUID，而debug（代码片段的最后一行）显示它是task.Context类。它是代码或文档中的错误吗？

谢谢。

Celery中“错误处理”回调的不同问题

0 个答案: