From: Nicholas D Steeves <sten@debian.org>
To: kreijack@inwind.it, Mark Harmstone <maharmstone@fb.com>,
linux-btrfs@vger.kernel.org
Cc: Mark Harmstone <maharmstone@meta.com>,
Qu Wenruo <quwenruo.btrfs@gmx.com>
Subject: RFC: mkfs.btrfs --subvols UI [was Re: [PATCH] btrfs-progs: add --subvol option to mkfs.btrfs]
Date: Fri, 02 Aug 2024 19:43:27 -0400 [thread overview]
Message-ID: <87plqqqyn4.fsf@digitalmercury.freeddns.org> (raw)
In-Reply-To: <eb3642fd-963b-4a14-9cd2-8339adcb58b7@libero.it>
[-- Attachment #1: Type: text/plain, Size: 1963 bytes --]
Hello,
Qu Wenruo <quwenruo.btrfs@gmx.com> writes:
> 在 2024/6/27 19:24, Mark Harmstone 写道:
>> From: Mark Harmstone <maharmstone@meta.com>
>>
>> This patch adds a --subvol option, which tells mkfs.btrfs to create the
>> specified directories as subvolumes.
>
> I have considered this feature in the past, but I do not have a good
> enough UI for that.
Recursive subvol backup & restore seems to be a desired feature, so I've
considered that in the following:
Goffredo Baroncelli <kreijack@libero.it> writes:
> On 27/06/2024 11.54, Mark Harmstone wrote:
>
> Could be possible to decouple the "--rootdir" and the "--subvol" options ?
> I.e. doing a first iteration where only the subvolume/subdir are created and a second one where
> all the subvolume are populated.
>
This also makes the most sense to me. In terms of UI, what are the
downsides to the following:
mkfs.btrfs --subvols 1 2 3 4 5 $device
# Creates a flat layout
mkfs.btrfs --subvols subdir/1 subdir/2 subdir/3 subdir/subsubdir/4 \
subdir/subsubdir/5 $device
# Creates subvols inside plain directories
unless this is specified:
mkdir.btrfs --subvols sv_cont0 sv_cont1 \
--subvols sv_cont0/1 sv_cont0/2 sv_cont1/3 sv_cont1/4 \
sv_cont2/5
This creates nested subvolumes, and "sv_cont2/5" will create subvol=5 in
a plain-dir named "sv_cont2" (the gotcha!/pitfall). Each subsequent
"--subvols foo bar baz" argument creates deeper level of nested
subvolumes. If the necessary parents haven't yet been created as
subvolumes, they are always created as directories.
And then "--rootdir" unpacks (yes, I'm conceptualising of rootdir as a
tarball). That said, I wonder if rootdir.img is a special
high-performance format, and wonder if it could benefit from a TOC to
collate subvolumes. This also seems orthogonal to a future btrfs send
image type would recurse and store subvol layout in a TOC of some kind.
Regards,
Nicholas
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 857 bytes --]
prev parent reply other threads:[~2024-08-02 23:43 UTC|newest]
Thread overview: 7+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-06-27 9:54 [PATCH] btrfs-progs: add --subvol option to mkfs.btrfs Mark Harmstone
2024-06-28 4:12 ` Qu Wenruo
2024-06-28 9:47 ` Mark Harmstone
2024-07-02 3:18 ` Qu Wenruo
2024-06-28 17:06 ` Goffredo Baroncelli
2024-07-01 11:34 ` Mark Harmstone
2024-08-02 23:43 ` Nicholas D Steeves [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=87plqqqyn4.fsf@digitalmercury.freeddns.org \
--to=sten@debian.org \
--cc=kreijack@inwind.it \
--cc=linux-btrfs@vger.kernel.org \
--cc=maharmstone@fb.com \
--cc=maharmstone@meta.com \
--cc=quwenruo.btrfs@gmx.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox