Currently, if for some reason the extraction of metadata for a resource fails, the adhoc task faildelay is triggered and the extraction task attempted again after a period of time.There is no cap on this and, as a result, a task could keep on going ad infinitum.
Recommend adding a hard cap, once the faildelay length is over this cap, the task should succeed, but a failed status is stored in the metadata_extractions table to record that extraction is not supported for the resource in question.
Currently, if for some reason the extraction of metadata for a resource fails, the adhoc task faildelay is triggered and the extraction task attempted again after a period of time.There is no cap on this and, as a result, a task could keep on going ad infinitum. Recommend adding a hard cap, once the faildelay length is over this cap, the task should succeed, but a failed status is stored in the metadata_extractions table to record that extraction is not supported for the resource in question.