So I already have glusterfs volume and it is working fine (I guess).
Now I am creating a new high-value prod setup and was starting to wonder - what is the "right" layout of nodes and bricks and/or replica count? I have about 9 nodes to spare. I was wondering how to use them to create a layout, that can sustain the fallout of two nodes/bricks. Namely, two of the nodes are actually VMs that are on the same host, so if I am to reboot or lose one of the hosts, two VMs are going out and I want that glusterfs volume to still be available and able to take reads and writes form it.
Please, any suggestions would be greatly appreciated!
