Open nclavaud opened 9 years ago
Just noticed a related PR (sorry): #40
Anything I could do to see this PR merged? Would help get rid of #40 also (duplicate).
Not sure about this. I do not like modifying a failed execution.
If we need something like this, I'm more for splitting up the Job entity into Job and JobExecution for example, where a Job can have many executions.
I do not like modifying a failed execution.
I get your point, but that's already what we've been doing here with automatic retries: https://github.com/schmittjoh/JMSJobQueueBundle/blob/master/Entity/Repository/JobRepository.php#L313
However, the Job / JobExecution distinction would indeed help keeping track of state changes and looks like a decent pattern.
Finally I find it, it wasn't happening only to me! +1 for this!
👍
Please accept this PR !
Hitting the
Retry
button for afailed
job will create a new job that does not reference the original. Even if the retry is successful, the failed job will remain infailed
state at the top of the board.The new job should reference the job that has failed previously. If the retry is successful, the original, failed job should reach state
finished
.