This is issue for master fail over:
Even master is down, mater is recovered, workers should also be recovered and reconnected to new master. But for last iteration of master, master down, worker should continue to run last iteration again. While all workers finish there last iteration and jump out loop in GuaguaWorkerService#run, mapper is also finished.
This is issue for master fail over:
Even master is down, mater is recovered, workers should also be recovered and reconnected to new master. But for last iteration of master, master down, worker should continue to run last iteration again. While all workers finish there last iteration and jump out loop in GuaguaWorkerService#run, mapper is also finished.