Current status: The half of "backup and restore" nobody likes to think about.
I lost my home PC's ZFS pool to a kernel panic-inducing bit of corruption. The lone disk in the pool has spent nearly 21,000 hours (2.4 years) of its 3-year life spinning rust, & SMART tests passed.
The kernel panic on `zpool import` means I basically lost everything on the disk.
But my USB backup disk is fine (I tested) & current enough. I should be able to install a new disk, install FreeBSD on it, & zfs send/recv my snapshots, I think.
Losses:
* An HDD apparently failing at 0.5 MTBF.
* This weekend.
* All data between final backup & failure: more or less 2 hours, the most consequential of which was mail.
Gains:
* Bigger HDD.
* XP in installing anything in FreeBSD's USB images.
* XP in restoring ZFS snapshots.
* Learning my backup procedure needs slight tweaking, but fundamentally everything was right.
Net:
I have now experienced ZFS disaster preparation, disaster, & disaster recovery!
* Searches "Seagate ST1000DM003 MTBF"
* Every seagate dotcom datasheet says or implies don't use it more than 2,400 hours per year. That's 6 hours/day, or 8-9/day if you skip weekends.
Yeah, this drive came free with cereal^W this PC, & I should've replaced it with a WD as soon as I got it. Oops.
Please make backup copies of your PC, folks, if you aren't already. Don't forget to test your backups. Don't forget to write down your PC recovery & backup restoring procedure. And don't forget to save that procedure somewhere you can get to if all your data gets wiped out.