AlmaLinux 10.2 Lavender Lion Beta supports older CPUs while RHEL moves on

https://fed.brid.gy/r/https://nerds.xyz/2026/05/almalinux-10-2-beta-lavender-lion/

Vanilla Arch with CachyOS repos layered on. Two kernels for two workloads. BTRFS so I can break things and not care.

Tested the rollback by installing KDE, hating it, restoring the snapshot. Two minutes, back to where I started. Did it again with GNOME.

https://blog.vintagetechie.com/posts/building-my-system-piece-by-piece

#Arch #BTRFS #Linux

Building My System, Piece by Piece · John Crenshaw

My career in backend development taught me how systems work from the inside out, but I spent those years using systems other people built. Now I get to ask what if we did it differently, and actually find out

John Crenshaw

For #beesd (#btrfs) is:

volume size * 4 / 1024, converted into MB, and then rounded up to the next whole integer that is divisible by 16 a good formula for estimating the hashTableSize?

Not quite sure what to think of that reply to my #Debian #ZFS package bug report yet... need to let that sink in.
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1135579
It's the first time I'm hearing that an "apt get dist-upgrade" on Debian Sid is supposed to be a "self-created problem". That a user should do x steps first. But maybe the rules for "contrib" packages are different?
Anyway, that makes me even more seriously consider switching from ZFS to #btrfs...
#1135579 - zfs-dkms does not compile with Linux 7.0 - Debian Bug report logs

Good gods I forgot @Ylfingr wrote this

https://git.brightfur.net/ylfingr/btrfs-unreachable-dependency-tracker

The whole thing is cursed. The whole SITUATION is cursed. I bet this is the issue I'm seeing again tonight.

*snootdesks.*

(btrfs can wind up in a state where you have to defrag _very specific files_ to free up the black-hole unreachable space, and defragging everything either does nothing or makes it WORSE.)

#btrfs

btrfs-unreachable-dependency-tracker

Recover from the creeping spread of UNREACHABLE

Sylvan Networks Git

Maybe I should really consider switching from #ZFS to #btrfs for my Linux/OpenWrt/Gluon dev partition. Deduplication + compression for their Git worktrees was the main reason why I added this partition in the first place, next to my ext4. And btrfs has these, too.

Don't quite remember why I chose #ZFS back then over #btrfs. Either the space savings were greater or the OpenWrt/Linux build times were faster in my tests.

#Btrfs copies read/write error statistics from a replaced disk in a RAID1 to the new one. One can reset those numbers manually.

Still, this is a bad design decision IMO.

#linux

Butterbian

Butterbian is a Debian 13 (Trixie) installer built around one idea: your system should be snapshotted before anything breaks it. Most distros leave BTRFS and Timeshift as an exercise for the user. Butterbian does it for you — subvolumes laid out correctly, Timeshift configured, grub-btrfs wired in, so you can roll back from GRUB if it all goes sideways.

Part of the @justaguylinux butter* ecosystem

https://butterbian.org/

#Linux #Debian #Butterbian #btrfs

Butterbian

Butterbian — Debian 13 with BTRFS, Timeshift snapshots, and grub-btrfs. Configured from day one.

Is there a way on #ZFS or #BTRFS to create a file that is the concatenation of two files?
So, cat a b > ab, but only taking up metadata?
Is the best way to just do the write and then ask the kernel to deduplicate it, and pray that it doesn't flush to disk in the meantime?
This would be really useful for serving smaller chunks of packages on #NixOS or #Guix, but also potentially on other distros as well, or even for general download optimization. (cc #bittorrent ?)

edit: Thanks, apparently there is an ioctl for this:
https://man7.org/linux/man-pages/man2/ioctl_ficlonerange.2.html

ioctl_ficlonerange(2) - Linux manual page

BTRFS Subvolumes On Debian Trixie - Easiest Tutorial

https://watch.linuxrenaissance.com/w/jqSKmBwtE5pkPqN41xKwzZ

BTRFS Subvolumes On Debian Trixie - Easiest Tutorial

PeerTube