1 |
On Wed, Mar 24, 2021 at 01:09:52PM -0400, Rich Freeman wrote: |
2 |
> On Wed, Mar 24, 2021 at 11:09 AM William Hubbs <williamh@g.o> wrote: |
3 |
> > |
4 |
> > On Wed, Mar 24, 2021 at 08:48:41AM +0100, Michał Górny wrote: |
5 |
> > > |
6 |
> > > What really can help is reflinking on filesystems supporting that. |
7 |
> > |
8 |
> > What really can help is more info instead of being terse like this. |
9 |
> > Which filesystems support it? |
10 |
> > |
11 |
> |
12 |
> According to Google right now: Btrfs, CIFS, NFS 4.2, OCFS2, overlayfs, and XFS |
13 |
> |
14 |
> Lizardfs ought to, but doesn't currently. zfs does not because clones |
15 |
> only are supported at the dataset level. |
16 |
> |
17 |
> In any case, if you're using coreutils cp to do the copy, just pass |
18 |
> --reflink=auto. Honestly, I have no idea why this isn't the default |
19 |
> behavior. Who wouldn't want instant copy operations that consume zero |
20 |
> space (aside from metdata)? If you're doing this in C or some other |
21 |
> language you would need to see if they have a library call to do it |
22 |
> easily - see man ioctl_ficlone. |
23 |
|
24 |
I'm using busybox, and I just checked and it also supports the |
25 |
--reflink=auto switch. |
26 |
|
27 |
I thought about coreutils, but with everything on the fs being moved |
28 |
around, I think that would get messy. |
29 |
|
30 |
Thanks a lot for the info Rich. :-) |
31 |
|
32 |
William |
33 |
|
34 |
> |
35 |
> -- |
36 |
> Rich |
37 |
> |