I'm integrating River as part of a bigger task system, and it's been great so far. But there's a particular missing feature that would be nice to have.
Currently, we are able to specify the maximum attempts of retries at both the worker and job level (https://riverqueue.com/docs/job-retries#limiting-retry-attempts):
_, err = riverClient.Insert(ctx, RetryOnceJobArgs{}, &river.InsertOpts{
// use a max attempts of 5 for this one job even though the job has a
// default limit of 1:
MaxAttempts: 5,
})On the other hand, the job retry delays can be set at the client and worker level (https://riverqueue.com/docs/job-retries#retry-delays), but not at the job level.
It would be great to be able to set the retry delay at job insertion, just like how the maximum attempt is.
A use-case for us is that users are able to create "rules" that control the dispatching of jobs to workers, and these rules can have their max attempts/delays customised.
Is this feature something that can be considered, or even feasible with the current River architecture?
I'm integrating River as part of a bigger task system, and it's been great so far. But there's a particular missing feature that would be nice to have.
Currently, we are able to specify the maximum attempts of retries at both the worker and job level (https://riverqueue.com/docs/job-retries#limiting-retry-attempts):
On the other hand, the job retry delays can be set at the client and worker level (https://riverqueue.com/docs/job-retries#retry-delays), but not at the job level.
It would be great to be able to set the retry delay at job insertion, just like how the maximum attempt is.
A use-case for us is that users are able to create "rules" that control the dispatching of jobs to workers, and these rules can have their max attempts/delays customised.
Is this feature something that can be considered, or even feasible with the current River architecture?