The failedJobRetryTimeCycle property can be used to define non static retry cycles using the ISO 8601 duration notation. For every duration in the list, the failed job will be retried once.
Example: "PT1M,PT5M,PT10M"
Will lead to one retry after 1 minute, 1 retry five minutes later and one retry 10 minutes later.
Requested feature:
Currently it is not possible to specify timer cycle definitions for failedJobRetryTimeCycle.
Example: "R2/PT1M,R1/PT5M,R/PT10M"
This would lead to two retries after 1 minute each, one retry 5 minutes later and MAX_INT retries after 10 minutes each.
This would help to create some kind of customizable retry backoff szenario where the amount of retries per time unit changes with time.