ETL with Hadoop - Big Data Made Easy: A Working Guide to the Complete Hadoop Toolset - page 316

Figure 10-36. Filter Columns step to eliminate unneeded columns
The next step is to sort the data and specify the order in which the columns will be sorted; Figure 10-37 shows a
compounded sorting key of manufacturer, model, and price.
Figure 10-37. Sort step to put columns in desired order
Finally, the data is stored via a results step, which specifies where the Map Reduce job will store its data and what
the field separator will be, as shown in Figure 10-38 . I use the default storage method of PigStorage and instruct the
data to be stored in the directory /data/talend/result/. Because this is a Map Reduce job, a successful outcome will
create a part file.
Figure 10-38. Results step provides for storage of the data