Originally Posted by Karl Beem
Then why string them together? Do you have files larger than 3TB?
I think this got touched above, but the big benefit is not needing to keep track of what drives are full/empty and what drive has what on it. Think of it like this, you can get big drives today, say 3TB that will hold a lot of stuff. So you put your pictures, movies, music on it. Works great, until you fill up that drive, so then you buy another drive, what do you do. Well without pooling you've got a couple options, one is you can copy say all your movies to the new drive leaving the pictures and music on the other. OK, not too bad, if you want your pictures or music it's on drive 1, movies drive 2. Not bad. Then you fill up drive 2 with movies, what now, get a new drive, and what, put all the new movies on the new drive? How long until you can't remember whether a given movie is "old" on drive 2 or "new" on drive 3.
Basically before too long, when you get a reasonably large number of drives, it becomes "impractical" to manually manage what goes on, and what lives on which drive. A pool makes it transparent to the user, everything is on the pool and you just go to the pool for everything.
Now granted, frontends like XBMC/MM/etc can pull media from multiple places so from a playback perspective you can hide that complexity, but you still have to deal with it when you add new media.
Originally Posted by ncarty97
Are you looking for data redundancy or drive pooling? For the data redundancy, neither setup is going to get you much over straight folder duplication. In the 1TB + 1TB + 2TB setup you'd have to use the 2TB as the parity drive, so you'd only have 2TB of drive space. With the 2TB + 3TB + 3TB, you'd be using one of the 3TB drives, leaving you 5TB of space, or 1TB extra over straight folder duplication.
Oh, I disagree completely, at least once you get past a drive or two. For any n drives worth of data, parity will save you n-1 drives. For 2 data drives that's not a big deal, 2-1 = 1 drive, but when you get up to more like 4 drives, that's 4-1 = 3 drives saved. I've got 6 data drives in my unraid so parity saves me 6-1 = 5 drives, that's not insignificant.
Another way to look at it is with folder duplication you lose 50% of your space to redundancy, with parity you lose n-1 drives, so for four drives you only use 25% for redundancy, for 8, it's only 12.5% (now it might be a little more when you start using different sized drives).
Originally Posted by ajamils
Don't really care about data redundancy as movies/music is not very critical plus I am not big fan wasting a drive for parity data (might change my mind if I get a good deal on drives
Just consider how long it will take you to re-rip/reorganize 1, 2, 3, 4TB of data if/when you lose a drive. It's that work that those of us that run redundant storage solutions are trying to protect. All my media is replacable from the original discs, but there's work involved in doing that. I value my time higher than the $100-200 for the extra drive required for redundancy.