the maxretry in level-jobs seems to be global, i.e. if anything in the history of the queue ever going to produce say maxretry number of errors the queue gets errored out. that's not cool at all. the correct way is have retry per worker. every worker gets that many retries then if at the end it still doesn't work then we do something about it.
the maxretry in
level-jobsseems to be global, i.e. if anything in the history of the queue ever going to produce say maxretry number of errors the queue gets errored out. that's not cool at all. the correct way is have retry per worker. every worker gets that many retries then if at the end it still doesn't work then we do something about it.