From FFS and ZFS to APFS: a comparative filesystem deep dive with a strong BSD angle
I wrote a long comparative article on filesystem design, but with a strong BSD spine running through it.
The piece covers FFS/FFS2, ZFS, BFS, NTFS, ext4 and APFS, with particular attention to:
- cylinder groups
- soft updates and crash consistency
- journaling vs CoW
- integrity models
- snapshots, clones and space sharing
- why APFS matters from an Apple/BSD lineage perspective
The APFS section is the destination, but I wanted the broader context first so the comparison would actually mean something.
Link:
Interested in where FreeBSD people think I was fair or unfair, especially on FFS/UFS and ZFS.