Btrfs is by design for enterprise use, but that doesn't stop normal home users from using it. By normal home user, I mean those people who don't use complicated setup like mirroring, or any level of RAID. They just use hard disks like hard disks, :-).
I know that if you use Btrfs then you don't need RAID because most of RAID features have already been builtin in Btrfs. My specific question is on Btrfs volume management -- I forgot wher I read, but I remember that you can add disks/partitions to Btrfs file system so that those small disks will looks like a huge logical big disk. Is that so? If so How?
Then, my question is, if I want to re-purpose a Btrfs partition for something else, how can I safely make sure that all files from under that particular partition have been copied out, so that the partition can be detached safely, provided that I've added another partitions with enough space to hold everything.
Ok, found the answer here, https://lwn.net/Articles/577961/
So let me move on to why I'm asking the question -- I've been using hard disks just like hard disks, no complicated setup like mirroring, or RAID or LVM. Previously, I've been partitioning my hard disks into small partitions, but having lots of small disk partitions is not fun, and it is a headache to maintain, so I'm considering to still make lots of small disk partitions but use them as the btrfs disk pools. What do you think the pros and cons of both approaches?
Now, the reason of having lots of small disk partitions is my ways of data-security. In case of hard disks errors, enterprises will just throw away that disk, whereas I will throw away that partitions with disk-errors. It may be frown upon by many people, but let me tell you, it works great. I have a lousy Seagate hard disk that keeps failing on me. I was forced to copy files out of the partition with disk-errors then stop using that partition. I think I've got three to six partitions with disk-errors already, but so far the rest are still fine. Having lots of small disk partitions makes it less painful to throw away those partitions with disk-errors.
So again, what do you think the pros and cons of above two approaches?