> Jean-Daniel Cryans <jdcry...@...> writes: > > > That can be done in a couple of seconds using the normal HBase client > in a multithreaded process, fed by a message queue if you feel like > it. What were you trying to achieve using MR? > > J-D
Yes, that's a point as well, to use or not to use the MapReduce. My intention was to use MapReduce to allow my task to be processed in parallel, where it is possible. I realize, that for a small amount of data, which is placed inside one online region in HBase, a regular Scan has got a much better performance, then a MapReduce routine. Still, looking forward, it's quite possible that the data will grow and reside in a number of online regions. This will be the situation, where the MapReduce will be more powerful. In other words, I just wanted to write code once and be prepared for the future. Andrey
