SSAS performance tuning (Case Study)

In the previous post, we have seen that only few smart moves/changes can make cube processing time far less than the current cube processing time. In this post, we will explore the way to identify the performance bottleneck and different ways to resolve it. In one of my project, cube processing was taking 8 hours where we were running 2 measure groups in parallel. Each measure group were having daily load partitions. Here are the initial stats: