Great Hive “WORDCOUNT” program example.
After working on the Map-Reduce word counting example and tinkering with a couple of other map-reduce programs, I thought I would try something with Hive. I ended up doing the same word count exercise because it is simple to set up, and though trivial, gives a nice contrast to using map-reduce directly.
Briefly, Hive presents a SQL like interface to Hadoop. The Hive query language is called QL and like SQL it is declarative. That means one would use QL to declare what results are wanted and not specify any detailed procedures for getting the result. Hive goes and figures out the best way to answer the request. This is in contrast to the map-reduce example I presented before which uses a procedural approach, one has to program all the steps in the form of mappers, and reducers in order to produce the desired result. QL hides the map-reduce procedures…
View original post 838 more words