nixos/zfs: disable redundant scheduler

[Artturin]

I (according to the comment) wrote this somewhere and since then it has spread to many configs.

https://github.com/search?q=artturin+ENV%7BID_FS_TYPE%7D%3D%3D%22zfs_member%22%2C+ATTR%7B..%2Fqueue%2Fscheduler%7D%3D%22none%22&type=code

#169457 (comment)

Does anybody know where I wrote it?

Description of changes

Things done

• Built on platform(s)
  • x86_64-linux
  • aarch64-linux
  • x86_64-darwin
  • aarch64-darwin
• For non-Linux: Is sandbox = true set in nix.conf? (See Nix manual)
• Tested, as applicable:
  • NixOS test(s) (look inside nixos/tests)
  • and/or package tests
  • or, for functions and "core" functionality, tests in lib/tests or pkgs/test
  • made sure NixOS tests are linked to the relevant packages
• Tested compilation of all packages that depend on this change using nix-shell -p nixpkgs-review --run "nixpkgs-review rev HEAD". Note: all changes have to be committed, also see nixpkgs-review usage
• Tested basic functionality of all binary files (usually in ./result/bin/)
• 23.11 Release Notes (or backporting 23.05 Release notes)
  • (Package updates) Added a release notes entry if the change is major or breaking
  • (Module updates) Added a release notes entry if the change is significant
  • (Module addition) Added a release notes entry if adding a new NixOS module
• Fits CONTRIBUTING.md.

[Artturin]

Thanks. The scheduler bit here was for some legacy functionality which was used to largely disable the IO scheduler since ZFS has it's own (only allow front and back merges). However, this functionality was removed a while ago since it was hard to support and users were encourages just to write a udev rule is they wanted to change the default scheduler.

openzfs/zfs#11838 (review)

openzfs/zfs@04344d5

It does not appear to affect the need to set this rule because people are still having the issue this PR fixes.

[Shawn8901]

I am using basically the same in my configs, so lgfm

[lopsided98]

The same person did claim that this should not be necessary anymore: openzfs/zfs#9609 (comment)

These days deadline, mq-deadline, noop, and none all perform well with ZFS, and users should not need to change the scheduler.

On the other hand, this doesn't appear to be always true in practice, and setting the default to none probably doesn't hurt. I just wanted to document upstream's advice.