I really like that you can specify what kind of anonymity level is needed for your workload, as well as how much disk space will be used under different circumstances (for example if 100% or 0%). You also have some nice options regarding compression in memory vs decompression from cache once it's already cached by other cores! It looks quite complex at first glance - but most people find their way around pretty quickly :) We've been using this software since 2014 which we're now running out of storage capacity with large amounts (>=500GB) per user/VM /app pool). The main problem here was finding an effective solution without having very expensive hardware requirements.