My three hosts are configured the exact same way physically:

2x250GB SSDs in Raid 1 for OS

2 x 2TB SSD in JBOD which I intend to use for Gluster (devices sda and sdb)

So on Step 3 (Volumes) of gluster deployment I'm leaving the volumes default (with no arbiter as I want replica 3).  So that is creating a volume for Engine, Data and VMstore

On step 4 (Bricks) I'm choosing jbod and next is where I'm getting confused.  I have to specify the device name and size for each volume.  It appears that I have to break up sda and sdb devices here.  I have 4 TB total between the two JBOD drives but does that mean I can only specify up to 2TB usable for the data volume (on either sda or sdb) then split the space on the second 2TB drive between engine and vmstore?  If so this seems like an awful waste of a 2TB drive for use for only engine and vmstore volumes.  With my JBOD configuration of two separate disks per hosts is there any way I can have a data volume use most of that space say 3.5TB and the rest goes to engine/vmstore?  It was my understanding that glusterFS would pool all available space together and I could use it however I wanted.  Can I achieve this without putting the 2TB disks in to a raid configuraiton?

Thanks!