Performance tuning and availability

Hi all,

Are there any experts here who can help me fine tune a FDB cluster? I need to be as fast as possible (obviously) with millions (possibly 1-2 billion) of writes a day and for it to have decent redundancy.

Version 6.3.23.

It is currently a 9 node cluster, 3 co-ordinators all running on r6i.xlarge (4 CPU / 32GB):

I know some of the config isnt right (ie storage option), but was hoping someone can give me some hints and tips to help make it fly!