ZFS 2.4.0 fixes SQLite issue - daemon may no longer be needed

[basnijholt]

Good news! 🎉

In #952 (locked so posting new issue), @ellie mentioned:

We will likely consider deprecating + removing it as soon as the issue with ZFS is resolved

The ZFS issue has now been resolved. PR openzfs/zfs#17584 by Rob Norris (Klara, Inc. / Wasabi Technology, Inc.) was merged and included in ZFS 2.4.0 (released 2025-12-18):

Commit	Description
fb7a8503bcfb	Linux: zfs_putpage: complete async page writeback immediately
a18c9edda621	Linux: sync: remove async/sync accounting
c3496b5cc6f8	Linux: zfs_putpage: document (and fix!) confusing sync/commit modes
7ac5440ecf53	ZTS: mmap_ftruncate test to confirm async writeback behaviour

The fix unlocks pages immediately once data is in the DMU/ZIL, rather than waiting for the ZIL itx callback. This eliminates the ftruncate() blocking on mapped page writeback that caused 500ms-5s delays at every shell prompt.

Root cause: openzfs/zfs#14290

Users on ZFS 2.4.0+ should no longer need:

• Atuin daemon mode
• sync=disabled workaround
• ext4 zvol workaround
• zfs_txg_timeout tuning

Is deprecating the daemon still on the table?

[ellie]

Excellent! Thank you for the heads up. I imagine it will take a while for it to make its way through all the repos

Is deprecating the daemon still on the table?

Yes, though I'd probably want to see if it's solving a different use case for people. It's been out for so long now that I wouldn't want to break any workflows

[mdekstrand]

It's been a while since I was poking at this, and didn't get it working satisfactorily but needed to move on before fully debugging, but the daemon also looked promising for dealing with kinda-slow network/cluster filesystems (such as the head node on my university's compute cluster). In such deployments, the usual caveats for running SQLite on a network file system aren't quite as applicable because you're usually only using an interactive shell on the head node.