Comment by kccqzy

I've only heard of this trick being employed on OLAP scenarios. Is this kind of partitioning also advisable for OLTP workloads?