Hi, We are running HBase for a while now with pre-splitted table and we end up having same number of regions on every node.
But as we add more hardware because of growing size we have very uneven cluster with different disk size, cpu, memory etc.. this causes problems related to disk utilization. So I was wondering is there a way to create a table in way where I have less regions on older hardware and more regions on newer hardware using pre-split strategy ? We are using hbase 0.98 with cdh 5.2.5. Thanks, Rahul
