If I want a progressive delay logic, i.e. incrementally increasing delay time, I need to do this myself in the worker when calling retry(). I might want to limit the retry by time (30 days) rather than count.
Given that, it doesn't make sense to me that the worker must worry about the producer not queuing up the job with the correct number of retries. I basically say max_retries=99999 now, which is ugly.
What I'd like to see is the retry() command gaining an option that allows users to skip the default max_retries logic, and allow the worker to decide if the retries have been exhausted.
If I want a progressive delay logic, i.e. incrementally increasing delay time, I need to do this myself in the worker when calling retry(). I might want to limit the retry by time (30 days) rather than count.
Given that, it doesn't make sense to me that the worker must worry about the producer not queuing up the job with the correct number of retries. I basically say max_retries=99999 now, which is ugly.
What I'd like to see is the retry() command gaining an option that allows users to skip the default max_retries logic, and allow the worker to decide if the retries have been exhausted.