Curious if there's an idiomatic way to handle this kind of use case:
I have a "retry" job that will run after a failure event is encountered (e.g., a remote server is down). It's important that this action take place eventually
I want this job to use exponential backoff (retryBackoff), but at some point I don't want to keep doubling the delay. For example, maybe after I hit a 15 min delay, I want to just keep retrying every 15 minutes (instead of going to 30 min, 60, 120, etc)
I'd like this job to just stick around until the failure is resolved (just use a very high value for expireInHours?)
I know that I could just implement the retry/backoff in my job handler, but is there a better way? Maybe using the onComplete callback to orchestrate? Or something else I'm overlooking?
Curious if there's an idiomatic way to handle this kind of use case:
retryBackoff
), but at some point I don't want to keep doubling the delay. For example, maybe after I hit a 15 min delay, I want to just keep retrying every 15 minutes (instead of going to 30 min, 60, 120, etc)expireInHours
?)I know that I could just implement the retry/backoff in my job handler, but is there a better way? Maybe using the
onComplete
callback to orchestrate? Or something else I'm overlooking?Thanks for your time.