So even if worker node fails between the job. It will resume the job?

And what about the data on the worker node?

Is it lost?