I have a question about layers on HDFS.
If i need to make subproducts, is better proccess with pig, spark or R the Virgin files and convert it into transformed files and insert in hive, o better attack the virgin files and show with any analitic sofware??
@Roberto Sancho You will be able to achieve much better performance by transforming files using simple peocessing in pig/Hive and create ORC Hive tables on transformed data.
View solution in original post