Moving a buried job back to a tube being watched by multiple workers causes same job to be reserved multiple times? about beanstalk_console HOT 4 CLOSED

So I realized that this problem still occurs even with just a single worker watching a tube (it just gets compounded with multiple workers). Looking at the code, I believe this occurs because of what's happening here:

https://github.com/ptrofimov/beanstalk_console/blob/1.7.7/lib/include.php#L765

If I bury a job immediately like in my sample code, it's possible for the do/while loop here to pick it up again ($job = $this->interface->_client->useTube($tube)->peekBuried();) before it is able to exit out via throwing a Pheanstalk_Exception_ServerException exception from not finding any jobs. That explains why I only see 1 buried job, despite it being executed and buried more than once. I'm guessing this a race condition type of problem?

from beanstalk_console.

How I see is:

• you have a worker listening for new jobs (and as you say bury them immediately)
• you click move ALL from Bury to Ready
• loop clicks in to peek all Bury
• identifies a buried N1, creates a copy puts to ready N2

•  workers takes the new copy N2 and buries as N2 
  

• loop deletes the buried N1 and loops again (since there is a new bury N2 already in place as the worker did created meanwhile the bury it keeps doing this)

I think you should pause the tube and then click the buttons. Otherwise because your workers are too fast, the console cannot do it's thing correctly.

from beanstalk_console.

That basically sums it up. I guess there are ways around it (like your suggestion of pausing) but that can be quite a hassle. If this is expected behavior, then this issue can be closed.

from beanstalk_console.

we cannot build around your logic, actually your logic of fast bury makes you this behaivior

from beanstalk_console.