FreeBSDs filesystem is one of the most reliable filesystems out there. It is old, but still performs pretty well and the age transfers to reliability. There ar

FreeBSDs filesystem is one of the most reliable filesystems out there. It is old, but still performs pretty well and the age transfers to reliability. There are 2 things creating problems now, though: first, it does not use journaling (by default) leading to expensive fsck times on big disks. Second, it could have faster meta-data operations. The filesystem has a construction such that a system crash can only lead to unreclaimed free space on the disk. If that is not the case, you are facing a hardware error. The primary role of fsck is to put back this free space. Journaling is now experimental in FreeBSD. ZFS is experimental in FreeBSD. As soon as these stabilizes, the problems mentioned here goes away entirely. Note that "experimental" means that the state is equivalent to most things linux: the crap actually works but have not yet seen the strains of time.