Retry failed jobs

[Douglas Simon Camargo]

Anyone have a example how to do a retry strategy with failed jobs using this lib, please ?

[Sudip Bhandari]

you can use maxFails from JobOptions.. by default it uses a default exponential backoff before another retry.. You can also supply a custom backoff function

work.JobOptions{
			Priority:       10,
			MaxFails:       5,
			MaxConcurrency: 100,
			//custom backoff function..
			Backoff: func(job *work.Job) int64 {
				return 600 + int64(math.Pow(10, float64(job.Fails+2)))
			},
		}

[Douglas Simon Camargo]

hmmm nice.

But my main question is: How do I call retry after an error in my code?

[Eric Alford]

@dscamargo you would just return the error object in the task handler. The library will retry all tasks that get an error returned assuming you have the retries configured.

[Douglas Simon Camargo]

Nice ! I will try.

Thank you @CodechCFA

[Gowtham Sai]

Hey folks,
I have a similar query but with a slight change.

Is there a way to tell the worker to retry without incrementing the retry attempted count?

[Sudip Bhandari]

@gwthm-in seeing this now but i was wondering what sort of use case you had in mind
(in case you didn't resolve already)

[Gowtham Sai]

@sudipidus thanks for getting back. This might not be pure work use case. But lemme try to describe what we are trying to achieve here.

We have an events system which pumps events from kakfa and schedules them to work. And in the work middleware we check the event order (sequential order lets say)., if the event is not in the sequence, we still want to process it, retry it (it's not a failure scenario). Hence the requirement is to retry, but not to increase retry failure count.