Hi all, There's one thing about Hadoop still not clear to me: are you supposed to treat the output files generated with the MapReduce tasks as input for other storage like database, index etc? Or are there applications that could use the files directly?
I'm experimenting with an application that could in a live situation generate tens of millions of lines in the MapReduce output files. I wonder if I could get a good performance out of storing them in relational db. I'd like to hear how other people are using the Hadoop MapReduce output files. -- Thanks and regards, Jeroen
