btrfs send on top level subvolumes that contain other subvolumes
(too old to reply)
john terragon
2014-10-19 15:00:30 UTC

Let's say I have a top-level subvolume /sub and that inside /sub I
have another subvolume say /sub/X/Y/subsub.

If I make a snapshot (both ro and rw give the same results) of /sub,
say /sub-snap, right now what I get is this

1) the /sub-snap/X/Y/subsub is present (and empty, and that's OK as
snapshot are not recursive) but it doesn't seem to be neither
a) an empty subvolume (because btrfs sub list doesn't list it)
b) a directory because, for example lsattr -d subsub gives me this
result "lsattr: Inappropriate ioctl for device While reading flags on

2) if /sub-snap is ro and I send it somewhere, then in the destination
sub-snap subsub is not present at all (which wouldn't be illogical,
given the non-recursive nature of snapshots).

So I'm wondering it all of this is the intended outcome when
snapshotting and sending a subvolume that has internally defined
subvolumes or if perhaps it's a bug.

I'm using kernel 3.17.1 patched for the recent ro snapshot corruption
bug and btrfs-progs from the 3.17.x branch in git.

To unsubscribe from this list: send the line "unsubscribe linux-btrfs" in
the body of a message to ***@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html