Note: running with Spark on Windows: copy bin/winutils.exe to some folder in local filesystem, like C:/hadoop/bin/winutils.exe create system varialbe HADOOP_HOME and set the value to 'hadoop' folder, like "C:\hadoop"