The typical reason for this issue is a library conflict crashing the Python shell. It's possible that a working cluster can run into this problem all of a sudden. This is because with a cluster restart libraries are fetched from the respective repositories. if there is a newer version
Identifying the problematic library is the first thing to resolve the issue. There is no easy way to identify the problematic library. It has to be a trial and error method. Uninstall each library, one at a time, and see if the issue persists.