Notes
Write-ups from my own systems: storage, backups and the things that went wrong with them.Each one links to the full post on negativeearth.net or tech.philcifone.com. If you read one, make it the post-mortem.
A daily ZFS status email
A cron job that emails me the health of every pool on every machine, with the latest snapshots, each morning. Daily rather than only on failure, so a missing email is itself a warning.
ZFS housekeeping
Shutting the whole homelab down to reorganize, then fixing what that turned up. A backup server that kept dropping its pools, a sanoid/syncoid setup that rolled backups back in time, and a new offsite copy through a Mac mini and Backblaze.
Escaping iCloud with a 44-year-old coreutil
Keeping full iPhone backups after cancelling iCloud. One symlink points macOS's local backups at their own ZFS dataset on an external drive.
NAS build and migration
Building a 12-bay NAS and moving the main pool and every service onto it. ZFS snapshots and send/receive copied the data block by block, so catching up the last changes took seconds instead of the hours rsync would have needed.
Storage pool notes
Working notes on my own pools, one page each, with the hardware, layout and datasets of each.
Post-mortem: a suspended ZFS pool
The morning status email said my main pool was suspended halfway through a scrub. What I checked, in what order, why a USB enclosure was the weak link, and how the pool came back with nothing lost.
AVIF instead of WebP
Why I moved my own sites' images to AVIF, and a small inotify and systemd watcher that converts new images as they arrive.