1024mb

Thank you, the only thing to be answered is point 1. One thing though, it seems I can't use cache pool with disk shares? EDIT: Yeah, I've searched and everywhere it's said that it's not possible

I want to preface that currently I'm using ZFS in stripe, so one drive, one pool on Debian. I haven't been able to properly test UnRAID being that I don't have spare drives right now, only one, so I'm not sure about the array and how it works. Now my questions: 1. I know that transfer speeds are decreased when using parity, will it be even further decreased if I use ZFS? Or will the speed be just like with any other filesystem? 2. When I add drives to an array, the whole array is displayed as a single logical drive or the drives are independently listed? I'm asking because I see a can share entire disks, but if the whole array is one single logical drive how is it that I can export individual drives? 3. About point 2. When I create a share and I select the array for storage, what happens? Let's say I have 2 drives in my array, I select the array and all disks for the share, does this means that a. UnRAID will distribute each single file across all the selected drives (i.e. If I copy some files, for each file X% will be on one drive and Y% on another) or b. UnRAID will distribute whole files across drives (i.e. I copy two files, one file will be on one drive and the other file on other drive)? 4. If I take a drive out of the array, will I be able to attach the drive to another server/computer or even an enclosure and browse the drive contents? (I'm guessing if point 3.a is how UnRAID works the answer to this question must be a No.) My system currently works in that I have every single ZFS drive mounted and I share the whole datasets (1 dataset per pool, so 1 drive 1 pool 1 dataset) on SMB and access each drive through my Windows machine as mounted network shares, so in Windows I have each server drive listed with their own drive letter. I would like to continue using this same system with UnRAID, is this possible?