I want to learn more about file systems from the practical point of view so I know what to expect, how to approach them and what experience positive or negative you had / have.
I found this wikipedia’s comparison but I want your hands-on views.
For now my mental list is
- NTFS - for some reason TVs on USB love these and also Windows + Linux can read and write this
- Ext4 - solid fs with journaling but Linux specific
- Btrfs - some modern fs with snapshot capability, Linux specific
- xfs - servers really like these as they are performant, Linux specific
- FAT32 - limited but recognizable everywhere
- exFAT - like FAT32 but less recognizable and less limited
I care a lot about filesystems.
BTRFS in FS-managed RAID configuration for automatic self-healing and snapshots for instant automated backups (though I keep a traditional backup too for protection against bugs and user error).
Storage is cheap compared to how much I value my data. BTRFS has very good support on Linux, integration with some backup tools, and I really want to use a FS that has full data checksums to make sure the data stays correct at rest. I like that I don’t have to use equal sized drives and can use whatever I have available, though I would appreciate a better read distribution model rather than the current where it just chooses a random drive to read from when multiple copies are available.
Disadvantages include difficulty accessing from Windows in my experience, less than stellar performance on HDDs, not very space efficient for small files systems because of the bulky metadata, and some uncommon RAID types don’t work correctly and will eat your data. I also don’t recommend it for use over USB because many such devices don’t correctly implement sync, and this is very important to stay on the correct transaction number and prevent file system inconsistencies. If I had to boot from USB, I wouldn’t pick BTRFS.
I don’t think exFAT or FAT32 offer POSIX permissions. I’m not sure if you could have a root file system there.
So, genuine question: why btrfs instead of zfs? Sounds like your use case would fit the latter.
BTRFS is zero effort on root, because it is included in kernel. ZFS on root is extra effort at least on Arch, due to licensing restrictions.
Additionally, at least for my use-case btrfs benefits me since it is less picky about drive sizes being the same and duplicating everything correctly - letting you essentially just throw additional storage at it as you acquire it.
Does ZFS handle data duplication on unequal sized volumes or heterogeneous pools? I don’t believe so, and BTRFS was a first class installer option.