django-q任务队列:失败的任务将永远保持处理。

Question

任何使用django-q任务调度程序的人：https://github.com/Koed00/django-q

(不是与数据库相关的Q库)

我有一大堆失败的任务。每次我运行django-q时，都会出现这些失败的任务：

18:06:14 [Q] INFO Process-1:9 processing [apart-sixteen-butter-friend]
18:06:14 [Q] INFO Process-1:10 processing [single-mississippi-bravo-fillet]
18:06:14 [Q] ERROR Failed [lithium-cola-batman-fanta] - No module named 'shootsta.user'
18:06:14 [Q] INFO Process-1:9 processing [ink-october-angel-california]
18:06:14 [Q] ERROR Failed [zulu-michigan-yankee-kilo] - No module named 'shootsta.user'
18:06:14 [Q] ERROR Failed [four-lemon-arizona-football] - No module named 'shootsta.user'
18:06:14 [Q] ERROR Failed [mississippi-fillet-winner-single] - No module named 'shootsta.user'
18:06:14 [Q] ERROR Failed [west-pennsylvania-asparagus-alabama] - 'BookingAnalytics' object has no attribute 'booking_uid'
18:06:14 [Q] ERROR Failed [wisconsin-pip-alanine-seventeen] - Can't switch from state 'job_complete' using method 'assign_camop'
18:06:14 [Q] ERROR Failed [finch-monkey-moon-oven] - 'AnalyticsFilter' object is not iterable
18:06:14 [Q] ERROR Failed [yellow-west-mango-papa] - 'AnalyticsFilter' object is not iterable
18:06:14 [Q] ERROR Failed [zulu-equal-mississippi-happy] - 'str' object has no attribute 'booking_uid'
18:06:14 [Q] ERROR Failed [six-beer-golf-blue] - 'str' object has no attribute 'booking_uid'
18:06:14 [Q] ERROR Failed [apart-sixteen-butter-friend] - 'str' object has no attribute 'booking_uid'
18:06:14 [Q] ERROR Failed [single-mississippi-bravo-fillet] - 'str' object has no attribute 'booking_uid'
18:06:14 [Q] ERROR Failed [ink-october-angel-california] - 'str' object has no attribute 'booking_uid'
18:06:15 [Q] ERROR reincarnated worker Process-1:4 after death
18:06:15 [Q] INFO Process-1:11 ready for work at 20244

这些失败任务中提到的相关问题已经解决。然而，失败的任务却不断出现。即使我从数据库中删除失败的条目(django_q_task表)，每次运行qcluster命令时都会重新创建它们。

如何处理这个问题并防止这些错误出现在输出中？

Answer 1

这是新添加的配置：attempt_count，您可以根据这个拉请求将其设置为1。

max_attempts不够好，而且令人困惑-- 因为最小重试是1 --因此，对失败任务的调用总数为2。

请记住，如果引发了'ack_failures': True，还需要设置Exception -以确认故障。

令人困惑的是，能起作用的是：

Q_CLUSTER = {
    'name': 'DjangORM',
    'workers': 1,
    'timeout': 600,
    'orm': 'default',
    'save_limit': 0,
    'ack_failures': True,
    'max_attempts': 1,
    'attempt_count': 1
}

Answer 2

我也有同样的问题--最简单的方法是通过Django管理web界面删除它们--只需登录并转到Django Q部分--删除失败的任务以及相关的调度和排队任务--应该是这样的。