Hello,

We are running a MapReduce program over HBase using TableMapper. But due to the limited amount of data ( which is sitting in one region server) , only 1 Map task is being spawned for the map phase. We are aware that one Map Task is spawned for every region server that contains keys in that table. But we want to have several Map tasks to increase parallelism. Can we manually either specify the number of keys in one region server or can we split the present region server ?
Thanks,
Naresh

Reply via email to