When Switch moves a job to the Problem jobs folder, it remembers the original location of the job in its internal job ticket. This makes it possible to move the job back to the location where it left the flow for another attempt at successful processing, without loosing any of the job's history (which is also stored in the internal job ticket). The process of moving the job back in the flow is called retrying the job.
It is meaningful to retry a job only after you have fixed the problem (or at least believe you did) by reconfiguring a process, adjusting the flow design, or tweaking the job itself.
Leave the job in the Problem jobs folder and leave its file name intact (including unique name prefix and file name extension); if you change the location or the file name Switch may loose the connection with the job's internal job ticket and it can no longer retry the job.
Switch offers various ways to retry a job: