Failure handling

• On failure, should increment count in DB + add failure date
• Retry doesn't happen for X amount of time(configurable)
• After Y number of retries(configurable), give up and mark row as unscrapeable
• Hit callback with that information