Small idea for Wiki page "Btrfs Maintenance"

Hope this is the right place to put it; I am not competent enough to write into the Wiki.

A few words maybe, whether caches can be troublesome, or can be always trusted, and the related btrfs check --clear-space-cache v1|v2 and --clear-ino-cache, and mounting with clear_cache?

1 Like

I only write what I have tested myself :wink:

  • I did install BTRFS on 30,000 devices
  • I did change ext4 partitions to btrfs
  • I did repaire BTRFS file systems (disk full)
  • I did rescued data from Btrfs (and never lost data)
  • I have never changed the default caches
  • I have never used btrfs check
  • I have never mounted with clear_cache

I would say:

  • whether caches can be problematic or are always trusted
    Don’t mess with Btrfs’ special options. They are best set by the team that maintains BTRFS in the kernel. In most cases, options do not require changes.

Tinkering with your filesystem is just as much fun :innocent: as sawing up the branch you’re sitting on. :see_no_evil:

If you want your system to work and not just for fun, rely on what the BTRFS programmers say. Many Manjaro users work with it. Others just enjoy it. :man_shrugging:

:footprints:

2 Likes

Not questioning your expertise – 30k devices is impressive. Guess btrfs has become robust enough for users like me, and we have a case of YMMV: I did the three bold things ( :crossed_fingers: , after backing up the partition :grinning:) and recovered from 177 unrecoverable errors. :fireworks: :sparkler:

Sheer luck maybe, but as a last resort? Left the defaults alone, though – here be dragons! :dragon: :dragon: :fire::dragon: :fearful: :cold_sweat: :dragon_face: :fire: